define lexical hypothesis

LEXICAL HYPOTHESIS

the theory that important natural characteristics and traits unique to individuals have become intrinsically embedded in our natural- language lexicon over time.

Leave a Reply

Your email address will not be published. Required fields are marked *

Latest Posts

A Quick Look at the History Behind Hypnosis

A Brief History of Brainwashing: The Science of Thought Control

A Deep Dive into the Social Psychology of Leadership

Counseling Approaches to Client Care: Theories to Apply in Practice

The Future Of Education: Can You Earn A Psychology Degree Online?

Insomnia & Mental Illness: What is the Correlation?

Stop Guessing: Here Are 3 Steps to Data-Driven Psychological Decisions

Getting Help with Grief: Understanding Therapy & How It Can Help

Exploring the Psychology of Risk and Reward

Understanding ADHD in Women: Symptoms, Treatment & Support

Meeting the Milestones: A Guide to Piaget's Child Developmental Stages

Counseling, Therapy, and Psychology: What Is The Difference?

Popular psychology terms, medical model, hypermnesia, affirmation, brainwashing, backup reinforcer, message-learning approach, affiliative behavior, confirmation bias.

The Lexical Hypothesis

Definition:

The Lexical Hypothesis, proposed by Gordon W. Allport and Henry S. Odbert in 1936, suggests that the most salient and important aspects of human personality are represented by the words found in natural language dictionaries.

Explanation:

According to the Lexical Hypothesis, individuals tend to describe and differentiate personality traits using words and phrases found in their linguistic systems. It assumes that if a characteristic or trait is significant enough to be recognized by a culture, it is likely to have a corresponding descriptor in the language of that culture.

Key Aspects:

The Lexical Hypothesis is built on the following key aspects:

• Synonym Frequency: Personality traits that are important to individuals are more likely to be addressed by multiple synonyms or descriptors in a given language. For example, the trait of “honesty” may also be captured by words like “truthfulness,” “integrity,” or “sincerity.”
• Importance and Visibility: Traits that hold greater significance and are more observable or noticeable tend to be represented by a larger number of words in a language. For instance, personality traits like “extraversion” or “conscientiousness” are typically described by numerous terms in dictionaries.
• Cross-Cultural Consistency: The Lexical Hypothesis assumes that universally recognized personality traits exist across cultures and languages. While the specific words used to describe these traits may differ, the underlying concepts are expected to be consistent.

Applications:

The Lexical Hypothesis has contributed significantly to personality psychology by guiding research on trait structure and taxonomy. It has aided in developing widely used trait models such as the Big Five personality traits (neuroticism, extraversion, openness, agreeableness, and conscientiousness). Researchers have extensively explored and validated lexical approaches to measure personality traits and identify their underlying factors.

The hypothesis has also informed various areas of applied psychology, including personnel selection, job fit assessments, clinical assessments, and relationship compatibility evaluations.

Conclusion:

The Lexical Hypothesis emphasizes the importance of language in understanding and describing human personality. It posits that the traits individuals consider significant and observable are reflected in the words and phrases they use. By leveraging natural language, researchers have gained valuable insights into trait structure, allowing for a better understanding of human personality and its implications in various areas of life.

• You are here
• Everything Explained.Today
• A-Z Contents
• Lexical hypothesis

Lexical hypothesis explained

In personality psychology, the lexical hypothesis [1] (also known as the fundamental lexical hypothesis , [2] lexical approach , [3] or sedimentation hypothesis ) generally includes two postulates :

1. Those personality characteristics that are important to a group of people will eventually become a part of that group's language. [4]

and that therefore:

2. More important personality characteristics are more likely to be encoded into language as a single word. [5] [6] [7]

With origins during the late 19th century, use of the lexical hypothesis began to flourish in English and German psychology during the early 20th century. [8] The lexical hypothesis is a major basis of the study of the Big Five personality traits , [9] the HEXACO model of personality structure [10] and the 16PF Questionnaire and has been used to study the structure of personality traits in a number of cultural and linguistic settings. [11]

Early estimates

Sir Francis Galton was one of the first scientists to apply the lexical hypothesis to the study of personality, stating:

Despite Galton's early ventures into the lexical study of personality, more than two decades passed before English-language scholars continued his work. A 1910 study by George E. Partridge listed approximately 750 English adjectives used to describe mental states, [12] while a 1926 study of Webster's New International Dictionary by M. L. Perkins provided an estimate of 3,000 such terms. [13] These early explorations and estimates were not limited to the English-speaking world, with philosopher and psychologist Ludwig Klages stating in 1929 that the German language contains approximately 4,000 words to describe inner states. [14]

Psycholexical studies

Allport & odbert.

Nearly half a century after Galton first investigated the lexical hypothesis, Franziska Baumgarten published the first psycholexical classification of personality-descriptive terms. Using dictionaries and characterology publications, Baumgarten identified 1,093 separate terms in the German language used for the description of personality and mental states. [15] Although this number is similar in size to the German and English estimates offered by earlier researchers, Gordon Allport and Henry S. Odbert revealed this to be a severe underestimate in a 1936 study. Similar to the earlier work of M. L. Perkins, they used Webster's New International Dictionary as their source. From this list of approximately 400,000 words, Allport and Odbert identified 17,953 unique terms used to describe personality or behavior.

This is one of the most influential psycholexical studies in the history of trait psychology . Not only was it the longest, most exhaustive list of personality-descriptive words at the time, it was also one of the earliest attempts at classifying English-language terms with the use of psychological principles. Using their list of nearly 18,000 terms, Allport and Odbert separated these into four categories or "columns":

Column I : This group contains 4,504 terms that describe or are related to personality traits. Being the most important of the four columns to Allport and Odbert and future psychologists, its terms most closely relate to those used by modern personality psychologists (e.g., aggressive, introverted, sociable). Allport and Odbert suggested that this column represented a minimum rather than final list of trait terms. Because of this, they recommended that other researchers consult the remaining three columns in their studies.

Column II : In contrast with the more stable disposition s described by terms in Column I, this group includes terms describing present states, attitudes , emotions, and moods (e.g., rejoicing, frantic). As a result of this emphasis of temporary states, present participles represent the majority of the 4,541 terms in Column II.

Column III : The largest of the four groups, Column III contains 5,226 words related to social evaluations of an individual person's character (e.g., worthy, insignificant). Unlike the previous two columns, this group does not refer to internal psychological attributes of a person. As such, Allport and Odbert acknowledged that Column III did not meet their definition of trait-related terms. Predating the person-situation debate by more than 30 years, [16] Allport and Odbert included this group to appease researchers of social psychology , sociology, and ethics.

Column IV : The last of Allport and Odbert's four columns contained 3,682 words. Termed the "miscellaneous column" by the authors, Column IV contains important personality-descriptive terms that did not seem appropriate for the other three columns. Allport and Odbert offered potential subgroups for terms describing behaviors (e.g., pampered, crazed), physical qualities associated with psychological traits (e.g., lean, roly-poly), and talents or abilities (e.g., gifted, prolific). However, they noted that these subdivisions were not necessarily accurate, as: (i) innumerable subgroups were possible, (ii) these subgroups would not incorporate all of the miscellaneous terms, and (iii) further editing might reveal that these terms could be used for the other three columns.

Allport and Odbert did not present these four columns as representing orthogonal concepts. Many of their nearly 18,000 terms could have been classified differently or put into multiple categories, particularly those in Columns I and II. Although the authors attempted to remedy this with the aid of three other editors, the average degree of agreement between these independent reviewers was approximately 47%. Noting that each outside reviewer seemed to have a preferred column, the authors decided to present the classifications performed by Odbert. Rather than try to rationalize this decision, Allport and Odbert presented the results of their study as somewhat arbitrary and unfinished.

Warren Norman

Throughout the 1940s, researchers such as Raymond Cattell [4] and Donald Fiske [17] used factor analysis to explore the more general structure of the trait terms in Allport and Odbert's Column I. Rather than rely on the factors obtained by these researchers, Warren Norman performed an independent analysis of Allport and Odbert's terms in 1963. [18] Despite finding a five-factor structure similar to Fiske's, Norman decided to use Allport and Odbert's original list to create a more precise and better-structured taxonomy of terms. [19] Using the 1961 edition of Webster's International Dictionary, Norman added relevant terms and removed those from Allport and Odbert's list that were no longer in use. This resulted in a source list of approximately 40,000 potential trait-descriptive terms. Using this list, Norman then removed terms that were deemed archaic or obsolete, solely evaluative, overly obscure, dialect-specific, loosely related to personality, and purely physical. By doing so, Norman reduced his original list to 2,797 unique trait-descriptive terms. Norman's work would eventually serve as the basis for Dean Peabody and Lewis Goldberg's explorations of the "Big Five" personality traits. [20] [21] [22]

Juri Apresjan and the Moscow Semantic School

During the 1970s, Juri Apresjan, a founder of the Moscow Semantic School, developed the systemic, or systematic, method of lexicography which utilizes the concept of the language picture of the world . This concept is also termed the naive picture of the world in order to stress the non-scientific description of the world which is found in natural language. [23] In his book "Systematic Lexicography", which was published in English in 2000, J.D.Apresjan puts forward the idea of building dictionaries on the basis of "reconstructing the so-called naive picture of the world, or the "world-view", underlying the partly universal and partly language specific pattern of conceptualizations inherent in any natural language". [24] In his opinion, the general world-view can be fragmented into different more local pictures of reality, such as naive geometry, naive physics, naive psychology, and so forth. In particular, one chapter of the book Apresjan allots to the description of lexicographic reconstruction of the language picture of the human being in the Russian language. [25] Later, Apresjan's work was the basis for Sergey Golubkov's further attempts to build "the language personality theory" [26] [27] [28] which would be different from other lexically-based personality theories (e.g. by Allport, Cattell, Eysenck, etc.) due to its meronomic (partonomic) nature versus the taxonomic nature of the previously mentioned personality theories. [29]

Psycholexical studies of values

In addition to research on personality, the psycholexical method has also been applied to the study of values in multiple languages, [30] [31] providing a contrast with theory-driven approaches such as Schwartz's Theory of Basic Human Values . [32] [33]

Similar concepts

Concepts similar to the lexical hypothesis are basic to ordinary language philosophy . [34] Similar to the use of the lexical hypothesis to understand personality, ordinary language philosophers propose that philosophical problems can be solved or better understood by an examination of everyday language. In his essay "A Plea for Excuses," J. L. Austin cited three main justifications for this method: words are tools, words are not only facts or objects, and commonly used words "embod[y] all the distinctions men have found worth drawing...we are using a sharpened awareness of words to sharpen our perception of, though not as the final arbiter of, the phenomena". [35]

Despite its widespread use for the study of personality, the lexical hypothesis has been challenged for a number of reasons. The following list describes some of the major critiques of the lexical hypothesis and personality models based on psycholexical studies. [7] [36]

• The use of verbal descriptors as material for analysis brings a pro-social bias of language into the resulting models. [37] [38] [39] Experiments using the lexical hypothesis indeed demonstrated that the use of lexical material skews the resulting dimensionality according to a sociability bias of language and a negativity bias of emotionality, grouping all evaluations around these two dimensions. [37] This means that the two largest dimensions in the Big Five model of personality (i.e., Extraversion and Neuroticism) might be just an artifact of the lexical method that this model employed.
• Many traits of psychological importance are too complex to be encoded into single terms or used in everyday language. [40] In fact, an entire text may be the only way to accurately capture and reflect some important personality characteristics. [41]
• Laypeople use personality-descriptive terms in an ambiguous manner. [42] Similarly, many of the terms used in psycholexical studies are too ambiguous to be useful in a psychological context. [43]
• The lexical hypothesis relies on terms that were not developed by experts. As such, any models developed with the lexical hypothesis represent lay perceptions rather than expert psychological knowledge.
• Language accounts for a minority of communication and is inadequate to describe much of human experience. [44]
• The mechanisms that resulted in the development of personality lexicons are poorly understood.
• Personality-descriptive terms change over time and differ in meaning across dialects, languages, and cultures.
• The methods used to test the lexical hypothesis are unscientific. [45]
• Personality-descriptive language is too general to be represented by a single word class , [46] yet psycholexical studies of personality largely rely on adjective s. [47]
• 16 Personality Factors
• Linguistic relativity

External links

• Language Personality Theory

Notes and References

• Book: Personality Theory . Oxford University Press . Crowne, D. P. . 2007 . Don Mills, ON, Canada . 978-0-19-542218-4.
• An alternative "description of personality": The Big-Five factor structure . Goldberg, L. R. . Journal of Personality and Social Psychology . December 1990 . 59 . 6 . 1216–1229 . 2283588 . 10.1037/0022-3514.59.6.1216. 9034636 .
• Book: The Psychology of Personality: Second Edition . Wiley-Blackwell . Carducci, B. J. . 2009 . Malden, MA . 978-1-4051-3635-8.
• Cattell . R.B. . The description of personality: basic traits resolved into clusters . Journal of Abnormal and Social Psychology . 1943 . 38 . 4 . 476–506 . 10.1037/h0054116.
• The lexical approach to personality: A historical review of trait taxonomic research . John, O. P. . Angleitner, A. . Ostendorf, F. . European Journal of Personality . 1988 . 2 . 3 . 171–203 . 10.1002/per.2410020302 . 15299845 . https://web.archive.org/web/20141026125224/https://pub.uni-bielefeld.de/luur/download?func=downloadFile&recordOId=1779427&fileOId=2312707 . 26 October 2014.
• Book: Miller . George A . The science of words . registration . 1996 . Scientific American Library . New York . 978-0-7167-5027-7.
• A defence of the lexical approach to the study of personality structure . Ashton, M. C. . Lee, K. . European Journal of Personality . 2004 . 19 . 5–24 . 10.1002/per.541 . 145576560 . https://web.archive.org/web/20120317220226/http://www.psy.uwa.edu.au/davidm/203/2007/lexical%20studies%20Ashton%20and%20Lee%202005b.pdf . 17 March 2012.
• Book: Personality: Determinants, Dynamics, and Potentials . Cambridge University Press . Caprara, G. V.. Cervone, D. . 2000 . New York . 978-0-521-58310-7.
• Goldberg. Lewis. The structure of phenotypic personality traits.. American Psychologist. 1993. 48. 1. 26–34. 10.1037/0003-066x.48.1.26. 8427480. 20595956.
• Ashton. Michael C.. Lee. Kibeom. Perugini. Marco. Szarota. Piotr. de Vries. Reinout E.. Di Blas. Lisa. Boies. Kathleen. De Raad. Boele. A Six-Factor Structure of Personality-Descriptive Adjectives: Solutions From Psycholexical Studies in Seven Languages.. Journal of Personality and Social Psychology. 86. 2. 2004. 356–366. 0022-3514. 10.1037/0022-3514.86.2.356. 14769090.
• Book: Handbook of Personality: Theory and Research, Third Edition . The Guilford Press . John, O. P.. Robins, R. W.. Pervin, L. A. . Paradigm Shift to the Integrative Big-Five Trait Taxonomy: History, Measurement, and Conceptual Issues . 2008 . New York . 114–158 . 978-1-59385-836-0.
• Book: An Outline of Individual Study . Sturgis & Walton . Partridge, G. E. . 1910 . New York . 106–111.
• The teaching of ideals and the development of the traits of character and personality . Perkins, M. L. . Proceedings of the Oklahoma Academy of Science . 1926 . 6 . 2 . 344–347 . 27 March 2012 . https://web.archive.org/web/20160303235502/http://digital.library.okstate.edu/OAS/oas_pdf/v06/p344_347.pdf . 3 March 2016 .
• Book: The Science of Character . George Allen & Unwin . Klages, L. . 1929 . London.
• Book: Trait-names: A psycho-lexical study. . Psychological Review Company . Allport, G. W.. Odbert, H. S. . 1936 . Albany, NY.
• The person-situation debate in historical and current perspective . Epstein, S.. O'Brien, E. J. . Psychological Bulletin . November 1985 . 98 . 3 . 513–537 . 10.1037/0033-2909.98.3.513 . 4080897.
• Consistency of the factorial structures of personality ratings from different sources . Fiske, D. W. . Journal of Abnormal and Social Psychology . July 1949 . 44 . 3 . 329–344 . 10.1037/h0057198. 18146776 .
• Toward an adequate taxonomy of personality attributes: Replicated factor structure in peer nomination personality ratings . Norman, W. T. . Journal of Abnormal and Social Psychology . June 1963 . 66 . 6 . 574–583 . 10.1037/h0040291 . 13938947.
• Book: 2800 personality trait descriptors: Normative operating characteristics for a university population . University of Michigan, Dept. of Psychology . Norman, W. T. . 1967 . Ann Arbor, MI .
• Book: Problems with Language Imprecision: New Directions for Methodology of Social and Behavioral Science . Jossey-Bass . Fiske, D. W. . Developing a taxonomy of trait-descriptive terms. . 1981 . San Francisco, CA . 43–65.
• Some determinants of factor structures from personality-trait descriptors . Peabody, D.. Goldberg, L. R. . Journal of Personality and Social Psychology . September 1989 . 57 . 3 . 552–567 . 10.1037/0022-3514.57.3.552 . 2778639. 19459419.
• An alternative "description of personality": The Big-Five factor structure . Goldberg, L. R. . Journal of Personality and Social Psychology . December 1990 . 59 . 6 . 1216–1229 . 10.1037/0022-3514.59.6.1216 . 2283588. 9034636 .
• Book: Apresi͡an, I͡Uriĭ Derenikovich. Systematic Lexicography . 2000. Oxford University Press. 978-0-19-823780-8.
• Web site: Apresjan J. (1992). Systemic Lexicography. In Euralex-92 Proceedings (part 1). p.4 . 2014-01-14.
• Book: Apresi͡an, I͡Uriĭ Derenikovich. Systematic Lexicography. https://books.google.com/books?id=WWPf0c024E8C. 2000. Oxford University Press. 978-0-19-823780-8. 101–143. The Picture of Man as Reconstructed from Linguistic Data: An Attempt at a Systematic Description.
• Web site: Golubkov's Language Personality Theory . webspace.ship.edu.
• Golubkov S.V. . 2002 . The language personality theory: An integrative approach to personality on the basis of its language phenomenology . Social Behavior and Personality . 30 . 6. 571–578 . 10.2224/sbp.2002.30.6.571 .
• Web site: PsychNews 5(1) . userpage.fu-berlin.de.
• Golubkov S.V. . 2002 . The language personality theory: An integrative approach to personalityon the basis of its language phenomenology . Social Behavior and Personality . 30 . 6. 573 . 10.2224/sbp.2002.30.6.571 .
• De Raad . Boele . Van Oudenhoven . Jan Pieter . A psycholexical study of virtues in the Dutch language, and relations between virtues and personality . European Journal of Personality . January 2011 . 25 . 1 . 43–52 . 10.1002/per.777. 227275239 .
• De Raad . Boele . Morales-Vives . Fabia . Barelds . Dick P. H. . Van Oudenhoven . Jan Pieter . Renner . Walter . Timmerman . Marieke E. . Values in a Cross-Cultural Triangle . Journal of Cross-Cultural Psychology . 28 July 2016 . 47 . 8 . 1053–1075 . 10.1177/0022022116659698. 151792686 .
• Schwartz . Shalom H. . Theory-Driven Versus Lexical Approaches to Value Structures . Journal of Cross-Cultural Psychology . 15 March 2017 . 48 . 3 . 439–443 . 10.1177/0022022117690452. 151514308 .
• De Raad . Boele . Timmerman . Marieke E. . Morales-Vives . Fabia . Renner . Walter . Barelds . Dick P. H. . Pieter Van Oudenhoven . Jan . The Psycho-Lexical Approach in Exploring the Field of Values . Journal of Cross-Cultural Psychology . 15 March 2017 . 48 . 3 . 444–451 . 10.1177/0022022117692677. 28502995 . 5414899 . free .
• Five big, Big Five issues: Rationale, content, structure, status, and crosscultural assessment . De Raad, B. . European Psychologist . June 1998 . 3 . 2 . 113–124 . 10.1027/1016-9040.3.2.113.
• Book: Philosophical Papers . Oxford University Press . Austin, J. L. . A Plea for Excuses . 1970 . London . 175–204.
• Book: Dumont, F.. A History of Personality Psychology: Theory, Science, and Research from Hellenism to the Twenty-first Century. Cambridge University Press. Trait theories and the psychology of individual differences.. 2010. New York. 149–182 . 978-0-521-11632-9.
• Trofimova . I. N. . 2014 . Observer bias: an interaction of temperament traits with biases in the semantic perception of lexical material. . PLOS ONE . 9 . 1 . e85677 . 2014PLoSO...985677T . 10.1371/journal.pone.0085677 . 3903487 . 24475048 . free.
• Trofimova. I.. Robbins. TW. Sulis. W.. Uher. J.. 2018. Taxonomies of psychological individual differences: biological perspectives on millennia-long challenges . Philosophical Transactions of the Royal Society, Biological Sciences. 373. 1744. 10.1098/rstb.2017.0152. 29483338 . 5832678 . free.
• Trofimova, I . etal . 2022. What's next for the neurobiology of temperament, personality and psychopathology? . Current Opinions in Behavioral Sciences. 45. 101143. 10.1016/j.cobeha.2022.101143. 248817462.
• A contrarian view of the five-factor approach to personality description. . Block, J. . Psychological Bulletin . March 1995 . 117 . 2 . 187–215 . 10.1037/0033-2909.117.2.187 . 7724687.
• Openness to experience: Expanding the boundaries of Factor V . McCrae, R. R. . European Journal of Personality . November 1994 . 8 . 4. 251–272. 10.1002/per.2410080404. 144576220 .
• A model and a method for uncovering the nomothetic from the idiographic: An alternative to the Five-Factor Model . Westen, D. . Journal of Research in Personality . September 1996. 30. 3. 400–413 . 10.1006/jrpe.1996.0028 . https://web.archive.org/web/20120507121610/http://www.psychsystems.net/Publications/1996/2.%20uncovering%20the%20nomothetic%20from%20the%20idiographic_Westen_jrnl%20of%20research%20in%20pers%201996.pdf . 7 May 2012. 10.1.1.523.5477 .
• Book: Personality Description in Ordinary Language . Wiley. Bromley, D. B. . 1977 . London . 978-0-471-99443-5 . registration .
• Book: Silent Messages . Wadsworth . Mehrabian, A. . 1971 . Belmont, CA . 978-0-534-00910-6 . registration .
• The Big Five versus nobody? . Shadel, W. G.. Cervone, D. . American Psychologist . December 1993 . 48. 12. 1300–1302 . 10.1037/0003-066x.48.12.1300.
• Personality-descriptive verbs . De Raad, B.. Mulder, E.. Kloosterman, K.. Hofstee, W. K. B. . European Journal of Personality . June 1988 . 2 . 2 . 81–96 . 10.1002/per.2410020204. 146758458.
• Eysenck . Hans Jurgen . The structure of phenoytypic personality traits: Comment . The American Psychologist . 1993 . 48 . 12 . 1299–1300 . 10.1037/0003-066x.48.12.1299.b.

This article is licensed under the GNU Free Documentation License . It uses material from the Wikipedia article " Lexical hypothesis ".

Except where otherwise indicated, Everything.Explained.Today is © Copyright 2009-2024, A B Cryer, All Rights Reserved. Cookie policy .

• May 31, 2022
• 10 min read

Personality 101: The Trait Approach & the Lexical Hypothesis

Human personality is a well-known concept in both academic and non-academic circles. This concept has raised the most diverse conclusions in both circles: from well-established factorial solutions to classifications of people based on what the Sorting Hat from Hogwarts would estimate. For a long time now, research in psychology has gained plenty of knowledge about human personality and its implications in everyday life, but this knowledge is either unknown or misunderstood by the general population. This situation calls for efforts to close this gap between what is known based on science and what is assumed to be true based on our random and subjective experience. The following 101 series explores the most consensual contemporary conceptualization of personality, the creative steps scientists took to arrive at this conceptualization using the lexical hypothesis, the specificities of this contemporary conceptualization by looking at each one of the Big Five traits, and the implications of each trait for individuals and societies.

The Personality 101 series is divided into 5 chapters:

Personality 101: The Trait Approach and The Lexical Hypothesis

Personality 101: Conscientiousness

Personality 101: Agreeableness and Extraversion

Personality 101: Emotional Stability

Personality 101: Openness to Experience

Thirty-one -year-old Robert often feels restless. He has problems sitting at a desk for more than a few minutes, cannot get organized, loses his keys and wallet, and forgets about his plans for the evening. He fails to achieve up to his potential at work. During the conversation, his mind wanders and he interrupts others, blurting out what he is thinking without considering the consequences. He gets into arguments. His mood swings and periodic outbursts make life difficult for those around him. Now his marriage is in trouble (John, 2021) .

This is a typical description of someone’s personality in a clinical setting. It conveys a good idea of how Robert is: disorganized with things and time causing evident consequences in his work, disorganized in speech, restless and impulsive. With this small paragraph, a clinician might already have an idea of what Robert is and what the goals would be in clinical intervention. However, something is lacking here. Surely, a short paragraph cannot summarize the wholeness of a human being, all their intricacies and idiosyncrasies; besides, the paragraph lacks some mention of the positive characteristics, too. Despite the risk of losing information, the task of trying to summarize someone's personality is useful, it helps with making life decisions easier: decisions like the search for a job, choosing a career, a partner, an adequate therapeutic procedure, and even one’s friends and hobbies.

What is needed is to summarize the personality of someone with the least loss of information possible, i.e. a descriptive model or a taxonomy of personality. “One of the central goals of scientific taxonomies is the definition of overarching domains within which large numbers of specific instances can be understood in a simplified way” (John, 2021, p. 38 ) . In plain words, a taxonomy about personality would be useful because it will allow the interpretation of the massive amount of information contained in one person’s behaviour using a very small group of categories. In this article, the best taxonomy for personality is going to be introduced along with the methodology followed to design it.

A taxonomy of human personality has been searched for a long time. Even in Ancient Greece Theophrastus would ponder: “why is that, while all Greece lies under the same sky and all the Greeks are educated alike, it has befallen us to have characters variously constituted?” (Theophrastus, 1909, p.77) . The most famous of the ancient attempts is Hippocrates' taxonomy in which he believed that different proportions of four bodily fluids or humour would manifest in the way people think, feel, and behave. A predominance of blood constituted a sanguine or social character, phlegm constituted a phlegmatic or easygoing character, black bile constituted a melancholic or analytical character, and yellow bile constituted a choleric or extraverted character (Chiao, 2018) .

Before going into more taxonomies of traits a note of recognition must be conceded to many other conceptualizations of personality that are not trait-based. Whereas the trait approach is one of the most frequently used nowadays, many authors in the history of psychology proposed models based on their own scientific and theoretical framework (Funder, 2012) . Freud’s framework, for instance, is based on the psychosexual development of the person, and he would argue that personality suffers many changes during childhood, also known as the stages of psychosexual development, but once adolescence arrives, personality becomes rather stable. Another key aspect of his theory is the division of personality into three components: the less conscious aspect, the id; the conscious experience of the person, the ego; and the social demands internalized in the individual, the super-ego (Funder, 2012) .

Later, other authors would propose different psychological processes as important components of human experience. Jung proposed the collective unconscious, the Anima and the Shadow as crucial mechanisms of the human psyche. More humanistic perspectives, like the ones championed by Abraham Maslow, Carl Rogers or Positive Psychology, focused on self-actualization processes, the acceptance of one's own experience, the hierarchies of motivations and needs, and the positive aspects of psychology like strengths and virtues (Funder, 2012) .

These are important contributions but some of them are elusive to scientific investigation, meaning that it is hard to apply the scientific method to answer the questions they pose. This is why, currently, they do not receive the same attention as the trait approach, and though there are some attempts to understand them, their significance is not as universal as the trait approach (Mastnak, 2021) .

The most basic tenet of the trait approach is, understandably, the trait. According to Allport (1931) , a trait has more than a nominal existence, is more than a generalized habit, is dynamic, it can be empirically or statistically found, is not unrelated to other traits, is not a moral quality, is not disproven if other behaviors appear in the behavioral repertoire of an individual, and it can be spotted in the individual and also in the population. Therefore, a trait is something real, can be found using statistical tools, and it is not disproven if other behaviors or emotions that go against this trait appear. This last feature of traits reveals an important nuance: a trait is the natural tendency of an individual, what the person would do almost spontaneously, the default mode of operating; this does not mean that, in a particular situation, an individual could not show behaviors, feelings, or thoughts that deviate from this natural tendency (Fleeson & Law, 2015) . Practically, this means that an extravert can sometimes act as an introvert, and vice versa.

Once the concept of trait has been cleared, let’s look at some of the attempts to define a taxonomy of personality in the modern history of Psychology. Raymond Cattell, a British-American psychologist, proposed 16 factors, or overarching categories obtained by means of statistical procedures, that comprise traits like Warmth or being outgoing and supportive, Social Assertiveness or being uninhibited and bold in social situations, Introversion or being reserved and clear-headed, and Independence or being self-sufficient (Cattell & Mead, 2008) . His theory led to the development of the Sixteen Personality Factor Questionnaire (16PF) that is still being used in vocational and educational settings. Another well-known theory of traits is the Eysencks’ theory of personality, in which there were only two bipolar factors accounting for all the variation in human personality: extroversion-introversion, and emotional stability-instability (Furnham et al., 2008) .

The methodology used by these authors to obtain such taxonomies is mostly based on the following process. The author would first get enough information about the topic by reading or gathering the conclusions obtained after years of experience in therapy and consultation. Once they think they have a solid theoretical framework from which to talk about personality, they will enumerate a set of traits that could explain and summarize humans across time and places. Although this is an oversimplification of the process, it serves one purpose: to show that, although it may be helpful for the patients of the author, it is not replicable or its replicability could be easily questioned. Therefore, since human personality is a universal phenomenon, a taxonomy that could replicate itself across contexts and individuals is needed (Mischel, 1996) .

There is a different process that overcomes the limitations of the previous one. This is the so-called lexical hypothesis, proposed by Galton (1949) , which states that every important human phenomenon must be somehow represented in the lexicon of a language, and since most languages are easily translated to others, the universality of the phenomenon could also be guaranteed. Based on this proposition, what authors would normally do is gather all the words used in a particular language for the phenomenon of interest from a representative sample of words in that language (some examples include dictionaries but also the transcripts of contemporary famous TV shows and movies), then they would ask a group of experts to analyze this list and determine which words are better at capturing the phenomenon under study (Ashton & Lee, 2007; Oreg et al., 2020; Parrigon et al., 2017) . This implies discarding synonyms and uncommon words from the sample. Later, they would approach a representative sample of individuals to categorize the phenomenon of interest (e.g. personality) according to the words established in the previous step. This will allow, by using proper psychometric and statistical techniques (i.e., factorial and multivariate statistical analyses), both to filter the best words that will be used and to create a refined measure of the phenomenon.

As of now, the scientific consensus is that the lexical hypothesis is the best solution found so far for the classification and understanding of personality (DeYoung et al., 2007) . In fact, one of the most famous and used taxonomies used at the moment, both in research, educational, clinical, and vocational settings, is the Big Five Taxonomy of personality, initially proposed by Costa & McCrae (1992) , and that have been further developed ever since. This taxonomy summarizes personality in five overarching traits: Conscientiousness, Agreeableness, Emotional Stability, Openness to Experience, and Extraversion. Each one of this will be further developed in the following posts of this 101 series.

One important note of warning is that, although this Big Five solution has proven to be very useful, the traits that conform it should not be understood as casual entities of human behaviors, instead they should be understood according to their real nature: a descriptive explanation of human personality. They do not explain casual relationships, they describe personality, they summarize it so that it can be easier to understand it (Fajkowska & Kreitler, 2018) . A second note is that these traits are not separable and completely distinguishable entities in and of themselves, they are correlated and there is some degree of overlap between some of them (Van der Linden et al., 2012) .

A final note, and the most important one, is that this taxonomy is based, as almost everything else in Psychology, on self-reports accounts. This is an important issue because, as Jung would say, “you are not what you say you’ll do, but what you do” (PsycholoGenie, 2014) . This fact poses a challenge, specially in personality research: are the tests really measuring psychological phenomena if they rely solely on self-report accounts? Are scientists not purposefully biasing their findings because self-reports suppose cheaper costs in research than observational or experimental reports (Galic et al., 2016; Olino & Klein, 2015) ? These questions are fueling some alternative research directions, like gathering behavioral data by means of wearables, cameras, and smartphones which, as of now, are both showing coincidences with the already extant research on the topic, and pushing its boundaries forward (Ihsan & Furnham, 2018) .

Personality psychology is an important and flourishing branch of Psychology. Its aim is to better understand human behaviors, thoughts, and emotions so that its knowledge would allow people to make better informed decisions in their life. Throughout history, personality has raised many questions to philosophers, writers, thinkers, and scientists and there have been many attempts to understand it. Though the validity of some of them could be recognized from a phenomenological point of view, the scientific method is not yet capable of working with them. Now, the consensus obtained in science is that the best solution found so far is the lexical hypothesis, as manifested in the relevance and extended use of the Big Five Theory of personality. These five traits have been proven useful in the description of many aspects of the human experience and their research is still a burgeoning theme in Psychology, though they are not free of limitations and improvements. Contemporary technologies are being gradually incorporated in the study of personality and their conclusions are solidifying the field and incorporating new findings. Only time will show how far the study of personality will get, and it is almost breathtaking to think that it all started with a man pondering why there were so many differences in people that were born under the same Ancient Greek sky.

Bibliographical References

Allport, G. W. (1931). What is a trait of personality? The Journal of Abnormal and Social Psychology , 25 (4), 368–372. https://doi.org/10.1037/h0075406

Ashton, M. C., & Lee, K. (2007). Empirical, Theoretical, and Practical Advantages of the HEXACO Model of Personality Structure. Personality and Social Psychology Review , 11 (2), 150–166. https://doi.org/10.1177/1088868306294907

Cattell, H. E. P., & Mead, A. D. (2008). The Sixteen Personality Factor Questionnaire (16PF). In The SAGE handbook of personality theory and assessment, Vol 2: Personality measurement and testing (pp. 135–159). Sage Publications, Inc. https://doi.org/10.4135/9781849200479.n7

Chiao, E. (2018, October 4). New study reveals four major personality types . The Johns Hopkins News-Letter. https://www.jhunewsletter.com/article/2018/10/new-study-reveals-four-major-personality-types

Costa, P. T., & McCrae, R. R. (1992). Four ways five factors are basic. Personality and Individual Differences , 13 (6), 653–665. https://doi.org/10.1016/0191-8869(92)90236-I

DeYoung, C. G., Quilty, L. C., Peterson, J. B., & nueva, E. a sitio externo E. enlace se abrirá en una ventana. (2007). Between facets and domains: 10 aspects of the Big Five. Journal of Personality and Social Psychology , 93 (5), 880–896. https://doi.org/10.1037/0022-3514.93.5.880

Fajkowska, M., & Kreitler, S. (2018). Status of the Trait Concept in Contemporary Personality Psychology: Are the Old Questions Still the Burning Questions? Journal of Personality , 86 (1), 5–11. https://doi.org/10.1111/jopy.12335

Fleeson, W., & Law, M. K. (2015). Trait enactments as density distributions: The role of actors, situations, and observers in explaining stability and variability. Journal of Personality and Social Psychology , 109 (6), 1090–1104. https://doi.org/10.1037/a0039517

Funder, D. C. (2012). The Personality Puzzle (pp. xxvii, 466). WW Norton & Co.

Furnham, A., Eysenck, S. B. G., & Saklofske, D. H. (2008). The Eysenck personality measures: Fifty years of scale development. In The SAGE handbook of personality theory and assessment, Vol 2: Personality measurement and testing (pp. 199–218). Sage Publications, Inc. https://doi.org/10.4135/9781849200479.n10

Galic, Z., Bubić, A., & Parmac Kovacic, M. (2016). Alternatives to self-reports: Conditional reasoning problems and IAT-based tasks. In The Wiley Handbook of Personality Assessment (p. 215.-227.). https://doi.org/10.1002/9781119173489.ch16

Galton, F. (1949). The Measurement of Character. In Readings in general psychology (pp. 435–444). Prentice-Hall, Inc. https://doi.org/10.1037/11352-058

Ihsan, Z., & Furnham, A. (2018). The new technologies in personality assessment: A review. Consulting Psychology Journal: Practice and Research , 70 (2), 147–166. https://doi.org/10.1037/cpb0000106

John, O. P. (2021). History, measurement, and conceptual elaboration of the Big‑Five trait taxonomy: The paradigm matures. In Handbook of personality: Theory and research, 4th ed (pp. 35–82). The Guilford Press.

Mastnak, W. (2021). Psychoanalysis and Qualitative Factor Analysis: A comparative meta- theoretical perspective . https://doi.org/10.13140/RG.2.2.11022.89922

Mischel, W. (1996). Personality and Assessment . Psychology Press. https://doi.org/10.4324/9780203763643

Olino, T. M., & Klein, D. N. (2015). Psychometric Comparison of Self- and Informant-Reports of Personality. Assessment , 22 (6), 655–664. https://doi.org/10.1177/1073191114567942

Oreg, S., Edwards, J. A., & Rauthmann, J. F. (2020). The situation six: Uncovering six basic dimensions of psychological situations from the Hebrew language. Journal of Personality and Social Psychology , 118 (4), 835–863. https://doi.org/10.1037/pspp0000280

Parrigon, S., Woo, S. E., Tay, L., & Wang, T. (2017). CAPTION-ing the situation: A lexically-derived taxonomy of psychological situation characteristics. Journal of Personality and Social Psychology , 112 (4), 642–681. https://doi.org/10.1037/pspp0000111

PsycholoGenie. (2014, August 12). A Comprehensive Collection of 60 Famous Quotes By Carl Jung. Psychologenie . https://psychologenie.com/collection-of-famous-quotes-by-carl-jung

Theophrastus. (1909). The characters of Theophrastus (R. C. Jebb, Trans. & J. E. Sandys, Ed.). London: Macmillan.

Van der Linden, D., Tsaousis, I., & Petrides, K. V. (2012). Overlap between General Factors of Personality in the Big Five, Giant Three, and trait emotional intelligence. Personality and Individual Differences , 53 (3), 175–179. https://doi.org/10.1016/j.paid.2012.03.001

David Saeteros

Arcadia has an extensive catalog of articles on everything from literature to science — all available for free! If you liked this article and would like to read more, subscribe below and click the “Read More” button to discover a world of unique content.

Let the posts come to you!

Thanks for submitting!

• Linguistic hypotheses
• Psychological theories
• Psychology terminology
• Philosophical theories
• Philosophical terminology
• Philosophy of language

Lexical hypothesis

Assessment | Biopsychology | Comparative | Cognitive | Developmental | Language | Individual differences | Personality | Philosophy | Social | Methods | Statistics | Clinical | Educational | Industrial | Professional items | World psychology |

Personality: Self concept · Personality testing · Theories · Mind-body problem

The Lexical Hypothesis [1] (also the Fundamental Lexical Hypothesis , [2] Lexical Approach , [3] or Sedimentation Hypothesis [4] ) is one of the most important and widely-used guiding scientific theories in personality psychology . [5] Despite some variation in its definition and application, the Lexical Hypothesis is generally defined by two postulates . The first states that those personality characteristics that are most important in peoples' lives will eventually become a part of their language. The second follows from the first, stating that more important personality characteristics are more likely to be encoded into language as a single word. [6] With origins in the late-19th century, use of the Lexical Hypothesis began to flourish in English and German psychology in the early 20th century. [4] The Lexical Hypothesis is the foundation for the HEXACO model of personality structure [7] and the 16PF Questionnaire and has been used to study the structure of personality traits in a number of cultural and linguistic settings. [8]

• 1.1 Early estimates
• 1.2.1 Allport & Odbert
• 1.2.2 Warren Norman
• 2.1 Philosophy
• 3 Criticism
• 5 References

History [ ]

Early estimates [ ].

Sir Francis Galton

Sir Francis Galton was one of the first scientists to apply the Lexical Hypothesis to the study of personality, [4] stating:

I tried to gain an idea of the number of the more conspicuous aspects of the character by counting in an appropriate dictionary the words used to express them... I examined many pages of its index here and there as samples of the whole, and estimated that it contained fully one thousand words expressive of character, each of which has a separate shade of meaning, while each shares a large part of its meaning with some of the rest. [9] :181 — Francis Galton ,  Measurement of Character , 1884

Despite Galton's early ventures into the lexical study of personality, over two decades passed before English-language scholars continued his work. A 1910 study by G. E. Partridge listed approximately 750 English adjectives used to describe mental states, [10] while a 1926 study of Webster's New International Dictionary by M. L. Perkins provided an estimate of 3,000 such terms. [11] These early explorations and estimates were not limited to the English-speaking world, with philosopher and psychologist Ludwig Klages stating in 1929 that the German language contains approximately 4,000 words to describe inner states. [12]

Psycholexical studies [ ]

Allport & odbert [ ].

Gordon Allport

Nearly half a century after Galton first investigated the Lexical Hypothesis, Franziska Baumgarten published the first psycholexical classification of personality-descriptive terms. Using dictionaries and characterology publications, Baumgarten identified 1,093 separate terms in the German language used in the description of personality and mental states. [13] Although this figure is similar in size to the German and English estimates offered by earlier researchers, Gordon Allport and Henry S. Odbert revealed this to be a severe underestimate in a 1936 study. Similar to the earlier work of M. L. Perkins , they used Webster's New International Dictionary as their source. From this list of approximately 400,000 words, Allport and Odbert identified 17,953 unique terms used to describe personality or behavior. [13]

This is one of the most influential psycholexical studies in the history of trait psychology . [4] Not only was it the longest, most exhaustive list of personality-descriptive words at the time, [4] it was also one of the earliest attempts at classifying English-language terms with the use of psychological principles. Using their list of nearly 18,000 terms, Allport and Odbert separated these into four categories or "columns": [13]

Allport and Odbert did not present these four columns as representing orthogonal concepts. Many of their nearly 18,000 terms could have been differently classified or placed into multiple categories, particularly those in Columns I and II. Although the authors attempted to remedy this with the aid of three outside editors, the average level of agreement between these independent reviewers was approximately 47%. Noting that each outside judge seemed to have a preferred column, the authors decided to present the classifications performed by Odbert. Rather than try to rationalize this decision, Allport and Odbert presented the results of their study as somewhat arbitrary and unfinished. [13]

Warren Norman [ ]

Throughout the 1940s, researchers such as Raymond Cattell [15] and Donald Fiske [16] used factor analysis to explore the overarching structure of the trait terms in Allport and Odbert's Column I. Rather than rely on the factors obtained by these researchers, [4] Warren Norman conducted an independent analysis of Allport and Odbert's terms in 1963. [17] Despite finding a five-factor structure similar to Fiske's, Norman decided to return to Allport and Odbert's original list to create a more precise and better-structured taxonomy of terms. [18] Using the 1961 edition of Webster's International Dictionary, Norman added relevant terms and removed those from Allport and Odbert's list that were no longer in use. This resulted in a source list of approximately 40,000 potential trait-descriptive terms. Using this list, Norman then removed terms that were deemed archaic or obsolete, solely evaluative, overly obscure, dialect-specific, loosely related to personality, and purely physical. By doing so, Norman reduced his original list to 2,797 unique trait-descriptive terms. [18] Norman's work would eventually serve as the basis for Dean Peabody and Lewis Goldberg's explorations of the Big Five personality traits . [19] [20] [21]

Similar concepts [ ]

Philosophy [ ].

Concepts similar to the lexical hypothesis are at the root of ordinary language philosophy . [22] Similar to the use of the Lexical Hypothesis to understand personality, ordinary language philosophers propose that philosophical problems can be solved or better understood through an exploration of everyday language. In his essay "A Plea for Excuses," J. L. Austin cited three main justifications for this approach: words are tools, words are not only facts or things, and commonly used words "embod[y] all the distinctions men have found worth drawing...we are using a sharpened awareness of words to sharpen our perception of, though not as the final arbiter of, the phenomena." [23] :182

Criticism [ ]

Despite its widespread use in the study of personality, the Lexical Hypothesis has been challenged for a number of reasons. The following list describes some of the major critiques levelled against the Lexical Hypothesis and personality models founded on psycholexical studies. [5] [6] [22] [24]

• Many traits of psychological importance are too complex to be encoded into single terms or used in everyday language. [25] In fact, an entire text may be the only way to accurately capture and reflect some important personality characteristics. [26]
• Laypeople use personality-descriptive terms in an ambiguous manner. [27] Similarly, many of the terms used in psycholexical studies are too ambiguous to be useful in a psychological context. [28]
• The Lexical Hypothesis relies on terms that were not developed by experts. [24] As such, any models developed with the Lexical Hypothesis reflect lay perceptions rather than expert psychological knowledge. [27]
• Language accounts for a minority of communication and is inadequate to describe much of human experience. [29]
• The mechanisms that led to the development of personality lexicons are poorly understood. [6]
• Personality-descriptive terms change over time and differ in meaning across dialects, languages, and cultures. [6]
• The methods used to test the Lexical Hypothesis are unscientific. [27] [30]
• Personality-descriptive language is too broad to be captured with a single word class , [31] yet psycholexical studies of personality largely rely on adjectives . [22]

See also [ ]

• Big Five personality traits
• 16 Personality Factors
• Trait theory
• Ordinary language philosophy

References [ ]

• ↑ Crowne, D. P. (2007). Personality Theory , Don Mills, ON, Canada: Oxford University Press.
• ↑ Goldberg, L. R. (December 1990). An alternative "description of personality": The Big-Five factor structure. Journal of Personality and Social Psychology 59 (6): 1216–1229.
• ↑ Carducci, B. J. (2009). The Psychology of Personality: Second Edition , Malden, MA: Wiley-Blackwell.
• ↑ 4.0 4.1 4.2 4.3 4.4 4.5 4.6 Caprara, G. V., & Cervone, D. (2000). Personality: Determinants, Dynamics, and Potentials , New York: Cambridge University Press.
• ↑ 5.0 5.1 Ashton, M. C., & Lee, K. (2004). A defence of the lexical approach to the study of personality structure . European Journal of Personality 19 : 5–24. Cite error: Invalid <ref> tag; name "defence" defined multiple times with different content
• ↑ 6.0 6.1 6.2 6.3 John, O. P., Angleitner, A., & Ostendorf, F. (1988). The lexical approach to personality: A historical review of trait taxonomic research . European Journal of Personality 2 : 171–203.
• ↑ (2004). A Six-Factor Structure of Personality-Descriptive Adjectives: Solutions From Psycholexical Studies in Seven Languages.. Journal of Personality and Social Psychology 86 (2): 356–366.
• ↑ John, O. P., Robins, R. W., & Pervin, L. A. (2008). Handbook of Personality: Theory and Research, Third Edition , 114–158, New York: The Guilford Press.
• ↑ Galton, F. (1884). Measurement of Character . Fortnightly Review 36 : 179–185.
• ↑ Partridge, G. E. (1910). An Outline of Individual Study , 106–111, New York: Sturgis & Walton.
• ↑ Perkins, M. L. (1926). The teaching of ideals and the development of the traits of character and personality . Proceedings of the Oklahoma Academy of Sciences 6 (2): 344–347.
• ↑ Klages, L. (1929). The Science of Character , London: George Allen & Unwin.
• ↑ 13.0 13.1 13.2 13.3 13.4 13.5 13.6 Allport, G. W., & Odbert, H. S. (1936). Trait-names: A psycho-lexical study. , Albany, NY: Psychological Review Company.
• ↑ Epstein, S., & O'Brien, E. J. (November 1985). The person-situation debate in historical and current perspective. Psychological Bulletin 98 (3): 513–537.
• ↑ Cattell, R. B. (October 1943). The description of personality: Basic traits resolved into clusters. Journal of Abnormal and Social Psychology 38 (4): 476–506.
• ↑ Fiske, D. W. (July 1949). Consistency of the factorial structures of personality ratings from different sources. Journal of Abnormal and Social Psychology 44 (3): 329–344.
• ↑ Norman, W. T. (June 1963). Toward an adequate taxonomy of personality attributes: Replicated factor structure in peer nomination personality ratings . Journal of Abnormal and Social Psychology 66 (6): 574–583.
• ↑ 18.0 18.1 Norman, W. T. (1967). 2800 personality trait descriptors: Normative operating characteristics for a university population , Ann Arbor, MI: University of Michigan, Dept. of Psychology.
• ↑ Fiske, D. W. (1981). Problems with Language Imprecision: New Directions for Methodology of Social and Behavioral Science , 43–65, San Francisco, CA: Jossey-Bass.
• ↑ Peabody, D., & Goldberg, L. R. (September 1989). Some determinants of factor structures from personality-trait descriptors. Journal of Personality and Social Psychology 57 (3): 552–567.
• ↑ 22.0 22.1 22.2 De Raad, B. (June 1998). Five big, Big Five issues: Rationale, content, structure, status, and crosscultural assessment. European Psychologist 3 (2): 113–124.
• ↑ Austin, J. L. (1970). Philosophical Papers , 175–204, London: Oxford University Press.
• ↑ 24.0 24.1 Dumont, F. (2010). A History of Personality Psychology: Theory, Science, and Research from Hellenism to the Twenty-first Century , 149–182, New York: Cambridge University Press.
• ↑ Block, J. (March 1995). A contrarian view of the five-factor approach to personality description.. Psychological Bulletin 117 (2): 187–215.
• ↑ McCrae, R. R. (November 1994). Openness to experience: Expanding the boundaries of Factor V. European Journal of Personality 8 (4): 251–272.
• ↑ 27.0 27.1 27.2 Westen, D. (September 1996). A model and a method for uncovering the nomothetic from the idiographic: An alternative to the Five-Factor Model . Journal of Research in Personality 30 (3): 400–413.
• ↑ Bromley, D. B. (1977). Personality Description in Ordinary Language , London: WIley.
• ↑ Mehrabian, A. (1971). Silent Messages , Belmont, CA: Wadsworth.
• ↑ Shadel, W. G., & Cervone, D. (December 1993). The Big Five versus nobody?. American Psychologist 48 (12): 1300–1302.
• ↑ De Raad, B., Mulder, E., Kloosterman, K., & Hofstee, W. K. B. (June 1988). Personality-descriptive verbs. European Journal of Personality 2 (2): 81–96.

• 1 Race and intelligence (test data)
• 2 Prostitution
• 3 Pregnancy fetishism

Find over 25,000 psychological definitions

lexical hypothesis

Browse dictionary by letter, psychology term of the day.

July 10th 2024

Lexical Hypothesis

The lexical hypothesis is a concept in personality psychology and psychometrics that proposes the personality traits and differences that are the most important and relevant to people eventually become a part of their language. It goes further to suggest that the most important concepts in personality become single descriptive words in a language. By using language as a resource and a sample a full spectrum and taxonomy of personality traits can be assembled. This concept has been around since the 1800s and many different methodologies have used the lexical hypothesis to develop personality taxonomies and lists. In 1936 Gordon Allport and Henry Odbert used the lexical hypothesis to conduct one of the most important and seminal studies in personality psychology. They used the dictionary to identify nearly 18,000 terms that described personality, behaviors, and traits. Raymond Cattell used computers in the 1940s to analyze Allport and Gordon’s terms and condensed them into 16 source traits or factors eventually developing the 16PF Personality Questionnaire which is still used to this day. Warren Norman reduced Allport and Gordon’s terms to 3,000 and eventually identified five overarching factors that encompassed most of human personality. Other researchers did the same and the Big Five (hyperlink ) personality factors became the most inclusive method of measuring personality. The Big Five are openness to new experience, conscientiousness, extraversion/introversion, agreeableness, and neuroticism. (OCEAN) and research has shown between the five 80% of personality variance can be accounted for.

Related terms:

• LEXICAL HYPOTHESIS
• Lexical Agraphia
• Dopamine (DA) Hypothesis
• MATURATION HYPOTHESIS
• CONTINUITY HYPOTHESIS
• RELATIONAL PRIMACY HYPOTHESIS
• PHYSICAL SYMBOL SYSTEM HYPOTHESIS
• PAIRING HYPOTHESIS
• NOVELTY HYPOTHESIS
• ALTERATION HYPOTHESIS

LEXICAL HYPOTHESIS

Lexical Hypothesis (LH) is a theory concerning the acquisition of language by children. It proposes that the primary factor in language acquisition is the availability of a large lexicon of words. This hypothesis is closely associated with the work of linguist Leonard Bloomfield, who proposed that language learning is based on the memorization of individual words and their meanings.

The LH posits that the primary factor in language acquisition is the availability of a large lexicon of words. This is based on the idea that children learn language through the process of memorizing the individual words and their corresponding meanings. The LH suggests that this process occurs through the child’s exposure to the language in their environment, with the child gradually acquiring an understanding of the language by memorizing the words and their meaning.

The LH has been supported by a number of studies which have shown that the availability of a large lexicon of words is an important factor in language acquisition. For example, a study by Clark and Clark (1977) found that children with larger vocabularies were more likely to acquire language faster than those with smaller vocabularies. Additionally, studies have shown that the rate of language acquisition is related to the amount of exposure to the language.

The LH has been criticized for its lack of specificity, as it does not explain the specific processes involved in language acquisition. Furthermore, it fails to account for the role of other factors, such as grammatical rules and the role of context in language acquisition. Additionally, some researchers have argued that the LH does not explain the process of language development beyond the initial acquisition stage.

Despite these criticisms, the LH continues to be an influential theory in language acquisition research. It provides an important framework for understanding how language is acquired and the role of vocabulary in the process. Additionally, it has provided researchers with an important basis for further research into the factors involved in language acquisition.

Clark, H. H., & Clark, E. V. (1977). Psychology and language: An introduction to psycholinguistics. Harcourt Brace Jovanovich.

Lenneberg, E. (1967). Biological foundations of language. New York, NY: Wiley.

Related terms

Logical error in rating, locomotor play, logo- (log-, longitudinal stability, loss of affect.

Measuring Lexical Quality: The Role of Spelling Ability

• Published: 13 April 2020
• Volume 52 , pages 2257–2282, ( 2020 )

Cite this article

• Sally Andrews 1 ,
• Aaron Veldre 1 &
• Indako E. Clarke 1  

6748 Accesses

2 Altmetric

Explore all metrics

The construct of ‘lexical quality’ (Perfetti Scientific Studies of Reading 11 , 357–383, 2007 ) is widely invoked in literature on word recognition and reading to refer to a systematic dimension of individual differences that predicts performance in a range of word identification and reading tasks in both developing readers and skilled adult populations. Many different approaches have been used to assess lexical quality, but few have captured the orthographic precision that is central to the construct. This paper describes, evaluates, and disseminates spelling dictation and spelling recognition tests that were developed to provide sensitive measures of the precision component of lexical quality in skilled college student readers – the population that has provided most of the benchmark data for models of word recognition and reading. Analyses are reported for 785 students who completed the spelling tests in conjunction with standardized measures of reading comprehension, vocabulary, and reading speed, of whom 107 also completed author recognition and phonemic decoding tests. Internal consistency analyses showed that both spelling tests were relatively unidimensional and displayed good internal consistency, although the recognition test contained too many easy items. Item-level analyses are included to provide the basis for further refinement of these instruments. The spelling tests were moderately correlated with the other measures of written language proficiency, but factor analyses revealed that they consistently defined a separate component, demonstrating that they tap a dimension of variability that is partially independent of variance in reading comprehension, speed, and vocabulary. These components appear to align with the precision and coherence dimensions of lexical quality.

Similar content being viewed by others

Precision in the measurement of lexical expertise: the selection of optimal items for a spelling assessment

Criteria for Evaluating Language Quality

Avoid common mistakes on your manuscript.

This paper has both practical and theoretical goals. The major practical goal is to validate and disseminate two measures of spelling ability that were developed to discriminate among samples of skilled, native English-speaking university students. Although other measures of spelling ability are available in batteries such as the Wide Range Achievement Tests (Wilkinson & Robertson, 2017 ), they were not specifically developed for university student populations. This may, in part, account for why spelling ability has received little attention in the literature on individual differences among skilled readers relative to measures of word and nonword reading (e.g., Kuperman & Van Dyke, 2011 ), reading speed (e.g., Jackson & McClelland, 1975 ), reading comprehension (e.g., Ashby, Rayner, & Clifton, 2005 ), and vocabulary (e.g., Yap, Balota, Sibley, & Ratcliff, 2012 ). The lack of attention to spelling ability also reflects a common assumption that, at least for skilled readers, measures of word identification and spelling tap the same dimension of individual variability, typically conceptualized as word identification or decoding . While spelling is seen as a useful index of reading development (e.g., Treiman, 2017 ), at higher levels of skill, reading and spelling words are often assumed to be 'two sides of the same coin' (Ehri, 2000 ), suggesting that spelling is unlikely to account for unique variance. More generally, the contribution of individual differences in word identification to explaining variability in reading comprehension is typically assumed to reduce across reading development relative to measures of comprehension-related factors like vocabulary and listening comprehension (e.g., Braze et al., 2016 ).

Challenging these assumptions, a series of studies using the tests of spelling dictation and spelling recognition evaluated in this paper has demonstrated that spelling ability accounts for unique variance in a variety of measures of performance across a range of single-word and sentence-processing tasks within samples of monolingual, English-speaking university students (see Andrews, 2008 , 2012 , 2015 ; Andrews & Veldre, 2019 , for reviews). Most of this research has been conducted in Andrews' laboratory at the University of Sydney, Australia, where the tests were developed, but they are now beginning to be used by independent psycholinguistic research groups (Adelman et al., 2014 ; Emmorey, Midgley, Kohen, Sevcikova Sehyr, & Holcomb, 2017 ; Eskenazi, Swischuk, Folk, & Abraham, 2018 ; Meade, Grainger, Midgley, Emmorey, & Holcomb, 2018 ; Slattery & Yates, 2018 ; Tan & Yap, 2016 ). It is therefore timely to report analyses evaluating the tests' internal consistency and validity and to make them publicly available along with details of the typical administration procedures and norms so that they can be consistently applied by other researchers to investigate individual differences among skilled readers.

The Lexical Quality Hypothesis

The theoretical goal is to evaluate the validity and utility of combining measures of spelling ability with other measures of written language proficiency to assess individual differences in lexical quality among skilled readers. Perfetti ( 1985 ) coined the term 'lexical quality' to refer to a critical determinant of the efficiency and effectiveness of the procedures involved in retrieving linguistic codes during reading comprehension. He subsequently refined the definition to refer to qualities of skilled readers' lexical knowledge (Perfetti & Hart, 2001 ): high-quality representations are "orthographically fully specified", represent redundant word-specific, context-sensitive phonology, and are "semantically more generalized and less context-bound" (Perfetti, 2007 , p. 359). This focus on the causal role of lexical knowledge distinguishes the lexical quality hypothesis (LQH) from many other accounts of individual differences in reading. Rather than attributing reading difficulties to deficits in phonological, semantic, or working memory processes , “the LQH is about knowledge that has not been acquired or practiced to a high-enough level” (Perfetti, 2007 , p. 380) to achieve the properties required for efficient, effective retrieval and subsequent higher-level processing.

The utility of the construct of lexical quality depends on how clearly it is defined. Perfetti ( 2007 ) emphasized the precision and flexibility of lexical knowledge. Precision is identified with the content of lexical knowledge: the specificity and completeness of the orthographic, phonological, grammatical, and semantic constituents of the lexical representation. Orthographic precision is particularly critical, because written forms become the gateway to lexical knowledge (Dehaene, 2009 ). Flexibility arises from the interconnectedness or binding between the different constituents of lexical representations. The precision and redundancy of high-quality representations strengthens the binding between the orthographic, phonological, and semantic constituents that define a word’s identity. These strong connections allow printed word forms to trigger synchronous, coherent activation of all components of the word’s identity required for the higher-order meaning integration processes underpinning effective comprehension (Perfetti, 2007 ). This property of greater coherence between the constituents of a lexical complex may be at least partially independent of the precision of the orthographic representation of a word (Andrews, 2015 ).

According to Perfetti’s definition, lexical quality is a graded, word-specific attribute. The quality of lexical representations varies within individuals as they gradually increase the size of their vocabulary and refine the specificity, redundancy, and interconnectedness of the knowledge stored for existing words through reading experience. There are also differences between individuals in the extent to which they have established high-quality representations for most of the words in their written vocabulary. Such differences could arise from genetic differences in foundational skills such as phonological awareness or orthographic learning (Byrne et al., 2008 ), or from environmental factors such as reading experience and methods of instruction. They may also interact with differences in reading strategy: readers who rely heavily on context to identify words may devote little attention to the details of words’ internal structure and therefore be less likely to develop fully specified orthographic codes for all words (Frith, 1986 ).

Assessing lexical quality

Perfetti's ( 2007 ) construct of lexical quality is increasingly widely invoked in the literature on both developing and skilled readers to refer to a systematic dimension of individual differences that predicts performance in a variety of word identification and reading tasks (e.g., Breadmore & Deacon, 2019 ; Rossi, Martin-Chang, & Ouellette, 2019 ; Slattery & Yates, 2018 ), but a range of different measures have been used to assess it. The most widely used indices of lexical quality are measures of vocabulary and decoding skill. Although word-level measures of decoding continue to predict significant variance in reading comprehension, even in skilled readers (e.g., Landi & Perfetti, 2007 ), vocabulary has been found to account for more variance than decoding at later stages of reading development (e.g., Protopapas, Sideris, Mouzaki, & Simos, 2007 ). In Verhoeven's systematic studies of the contribution of lexical quality to the development of reading comprehension in late-primary-age Dutch children, tests of both the breadth and depth of vocabulary accounted for a substantial proportion of variance in reading comprehension in a large cross-sectional sample, after controlling for decoding measures of word and nonword reading, short-term memory, and nonverbal intelligence (Swart et al., 2017 ). Vocabulary and category fluency also predicted growth in reading comprehension in a longitudinal study over grades 4 to 6 (Nouwens, Groen, Kleeman, & Verhoeven, 2017 ). Measures of 'lexical richness' obtained from tests of synonym knowledge and verbal analogies have also been found to predict eye-movement measures of lexical processing and semantic integration in English-speaking adolescents' sentence reading (Luke, Henderson, & Ferreira, 2015 ).

Individual differences in vocabulary, decoding, and meaning retrieval have also been used to assess lexical quality in adult readers. Many of the investigations conducted by Perfetti's group have used scores derived from factor analyses of a broad battery of tests of reading-related skills administered to large samples of college-student readers to define "functionally distinct dimensions of variability" (Taylor & Perfetti, 2016 ). Early studies (e.g., Perfetti & Hart, 2002 ; Landi & Perfetti, 2007 ) identified separate 'meaning knowledge' (assessed by vocabulary and reading comprehension) and 'form knowledge' factors (defined by spelling, phonology, and decoding) and suggested that form knowledge split into separate orthographic and phonological factors among less-skilled readers (Perfetti & Hart, 2002 ), supporting the LQH's assumption that lexical constituents become more integrated at higher levels of reading skill. Taylor and Perfetti’s ( 2016 ) more recent research using an expanded battery including subjective reports of reading history (Lefly & Pennington, 2000 ) and a measure of print exposure based on author recognition (Stanovich & West, 1989 ) found that the strongest source of shared variance was a 'reading experience' factor defined by measures of reading speed, print exposure, book reading, and reading attitude, which was independent of a robust 'lexical knowledge' factor defined by decoding, word recognition, and spelling skill. Both factors predicted more efficient eye movements – more word skipping and fewer regressions – during reading of short passages, but the lexical factor was a stronger predictor of these reading behaviors for texts containing unfamiliar, recently acquired words. Convergent evidence that word identification predicts individual differences in eye movements in a broader community sample of adults derives from Kuperman and Van Dyke's ( 2011 ) finding that word identification and rapid naming were the only measures from a large battery of tests assessing decoding, working memory, and listening and reading comprehension that significantly predicted unique variance in eye movements during sentence reading. Better performance on these measures was associated with efficient oculomotor control that Kuperman and Van Dyke suggested as being "an affordance of the overall quality of … [the reader's] lexical representations" (p. 56).

The test-battery approach adopted in the studies described above is very resource-intensive, particularly for the large samples that are desirable for individual-differences research. Many researchers have therefore attempted to capture individual differences in lexical proficiency with a single or limited set of measures. Vocabulary has been systematically used as a coarse index of 'lexical integrity' in Yap and colleagues’ thorough investigations of how individual differences modulate both behavioral performance and the underlying processes revealed by mathematical modeling methods across a range of word identification tasks (e.g., Pexman & Yap, 2018 ; Yap et al., 2012 ; Yap, Tse, & Balota, 2009 ). Another relatively widely used single measure that aims to assess the reading experience assumed to underpin lexical quality is the Author Recognition Test (ART). Originally developed by Stanovich and West ( 1989 ), and subsequently refined and extended by Acheson, Wells, and MacDonald ( 2008 ), these quick, easily administered tests use the ability to discriminate between real and fabricated authors (or magazine titles; Acheson et al., 2008 ) as a surrogate measure of the extent of exposure to print (see Moore & Gordon, 2015 , for a review).

The accumulating evidence that a range of measures of lexical proficiency predict systematic variance in skilled readers' performance not only in single-word identification and priming tasks but also in eye-movement measures of sentence reading confirms that individual differences in word-level processes remain significant predictors of variance among adult readers. However, to provide evidence for a specific contribution of lexical quality requires specification of how the predictor variables map to the precision, redundancy, and coherence of lexical representations that define their quality. As highlighted by Braze, Tabor, Shankweiler, and Mencl ( 2007 ), vocabulary knowledge is central to general linguistic comprehension processes that are shared between spoken and written language processing. Correlations between vocabulary and reading performance may, therefore, tap general comprehension processes rather than the quality of reading-specific lexical representations. Consistent with this interpretation, factor analyses typically show that vocabulary loads on the same latent factor as listening comprehension (Braze et al., 2016 ). However, in two independent studies of large community samples of adult readers, vocabulary also accounted for a small, but significant, component of unique variance in reading comprehension when word decoding and listening comprehension were controlled for (Braze et al., 2007 ; Braze et al., 2016 ), suggesting a reading-specific contribution. Braze et al. ( 2016 ) suggested this may be because higher-quality representations "that incorporate subtle gradations of meaning, may integrate more flexibly into representations of discourse of narrative and … be more readily recognized in context" (p. 447). Nevertheless, in studies that only measure vocabulary, it is not clear whether observed effects reflect lexical quality specifically, or factors related to general comprehension – or some combination of the two. Similarly, ART measures of print exposure are typically at least moderately correlated with measures of vocabulary and word identification (e.g., Moore & Gordon, 2015 ). Moreover, as well as contributing to the refinement of lexical representations, the reading experience that these tests are presumed to assess will influence a range of reading processes (Falkauskas & Kuperman, 2015 ).

Orthographic precision and lexical quality

Precision is central to Perfetti's ( 2007 ) definition of lexical quality, but little attention has been paid to establishing the extent to which the measures of individual differences used in studies of skilled reading capture this attribute. Orthographic precision is particularly critical to lexical quality: successful word identification requires readers to extract the relevant features from the perceptual input and map them to existing lexical representations. Phonology shapes the orthographic units that need to be extracted, but neuroimaging evidence suggests that skilled readers develop a specialized visual system for mapping visual input to word-specific knowledge (Dehaene et al., 2010 ). Spelling ability provides a direct index of the orthographic precision of readers' lexical representations of known words that the LQH suggests may play a specific role in predicting effective reading, independently of other measures of lexical proficiency.

Consistent with this view, recent studies of skilled readers that have indexed lexical quality by combining the spelling dictation and recognition tests described in this paper with standardized measures of vocabulary and reading comprehension have demonstrated that spelling ability predicts unique variance both in masked priming studies of single-word identification tasks (Andrews & Hersch, 2010 ; Andrews & Lo, 2012 , 2013 ; Andrews, Lo, & Xia, 2017 ) and in behavioral (Andrews, 2008 ; Hersch & Andrews, 2012 ) and eye-movement indices of sentence reading (Drieghe, Veldre, Fitzsimmons, Ashby, & Andrews, 2019 ; Veldre & Andrews, 2014 , 2015a , 2015b ; Veldre & Andrews, 2016a , 2016b ; Veldre, Drieghe, & Andrews, 2017 ). Independent studies using one or both of the same spelling tests have provided converging evidence for the unique contribution of spelling ability to predicting adult readers’ lexical decision performance (Adelman et al., 2014 ), eye movements during sentence reading (Slattery & Yates, 2018 ), and electrophysiological indices of word processing (Meade et al., 2018 ). A second electrophysiological study found that the selective effect of spelling was more marked in pre-lingually deaf adults than hearing readers (Emmorey et al., 2017 ), suggesting that orthographic precision may play a particularly important role in reading development when phonological processes are compromised.

This accumulated evidence confirms that tests of spelling ability account for individual differences among skilled adult readers that are not captured by other measures of written language proficiency. However, further evidence about the internal consistency and validity of these tests is required to confirm and elaborate the relationship between spelling and lexical quality. To address these issues, the present paper reports analyses of data collated from over 800 individuals tested across nine independent samples of between 46 and 110 University of Sydney students who completed individual-differences tests in conjunction with their participation in eye-tracking studies of sentence reading.

The present research

The effects of spelling ability summarized above were derived from tests of spelling dictation and spelling recognition developed to discriminate among samples of skilled university student readers that have been briefly described in a number of previous papers (e.g., Andrews, 2008 , 2012 , 2015 ), and in greatest detail by Andrews and Hersch ( 2010 ). Spelling production tasks, like dictation, directly test the precision of orthographic knowledge, but some theories of spelling assume that spelling production relies on different information from that required for recognition of correct spellings. Readers may rely on partial orthographic information to identify words (e.g., Frith, 1980 ) and therefore be able to correctly recognize words for which they cannot produce a correct spelling. Performance on spelling recognition tasks is also influenced by the type of misspellings included. Katz and Frost ( 2001 ) found that participants were more likely to accept a repeated misspelling as being correct when it was phonologically plausible, even if it had been correctly rejected on its first presentation, and adopted laxer criteria for judging spelling acceptability when phonologically implausible spellings were included in the recognition list than when all misspellings were phonologically plausible. Such findings demonstrate the role of decision processes in spelling recognition tasks and suggest that participants can be induced to vary the relative weighting of orthographic and phonological information. More extreme differences between production and recognition tasks are suggested by cognitive neuropsychological evidence of dissociations between reading and spelling performance in brain-injured patients that has been interpreted as indicating separate input and output representations that can be independently accessed and damaged (e.g., Ellis, 1993 ). Such views predict that readers may be able to correctly recognize the spelling of words that they cannot accurately produce by relying on their more accurate input representation (Holmes & Babauta, 2005 ).

Thus, spelling recognition tasks may tap factors not captured by spelling dictation. To comprehensively assess spelling ability it is therefore important to include both measures. As reported by Andrews and Hersch ( 2010 ), the spelling dictation and recognition tests both demonstrate high test–retest reliability ( r  = 0.90 and 0.93, respectively), but they have not previously been analyzed for internal consistency. This is the goal of the first set of analyses reported here.

The further goal of the present research is to extend the evidence for the convergent and divergent validity of the spelling tests by assessing their relationships with other measures of written language proficiency across a large sample of participants. Before tackling this question, we addressed an issue relevant to assessing reading comprehension and vocabulary, two of the other major indices of lexical quality. Specifically, we investigated the implications of varying the time limits allowed for the reading comprehension and vocabulary subtests of the Nelson-Denny Reading Test (NDT; Brown, Fishco, & Hanna, 1993 ), a widely used standardized measure of adults’ reading proficiency. This test was developed for use with students from grade 9 of high school through to the fourth year of college/university. Our extensive experience administering the test to samples of university students has revealed that a substantial number of participants complete one or both subtests before the recommended time limit expires. Presumably because of the potential reduction in the tests’ discriminative power, combined with pragmatic constraints on testing time, applications of these tests in Perfetti’s studies of individual differences in lexical quality among skilled readers have typically reduced the time limits for each subtest to half the length required for standard administration (e.g., Perfetti & Hart, 2001 ; Taylor & Perfetti, 2016 ). The same reduced time limits have been used in approximately half of the studies collated for the present research, while the remainder used the standard time limits. This provided an opportunity to evaluate the impact of administration time on the distribution of NDT scores and their relationship to scores on other tests.

Participants

Data were collated for 813 students who participated in exchange for credit in introductory Psychology courses. The mean age of the participants was 19.59 years Footnote 1 (SD = 3.65 years), and approximately 71.3% of the sample was female. The recruitment criteria specified that participants spoke English, and began learning to read and write English by no later than age 6. Most participants (87.1%) reported that English was the first language they learned to speak, and English was the first language that virtually all (96.6%) the sample learned to read and write.

The individual-differences data were collected over a period of approximately 5.5 years (late 2011–early 2017) as part of a series of nine eye-movement experiments investigating the role of written language proficiency in sentence reading. Almost all participants completed the two spelling tests (798 participants). In addition, most participants also completed the vocabulary, reading comprehension, and reading rate subsections of the NDT (785 participants). The Author Recognition Test (ART; Moore & Gordon, 2015 ) and the phonemic decoding subtest of the Test of Word Reading Efficiency (TOWRE; Torgesen, Wagner & Rashotte, 1999 ) were completed by 107 participants. The individual-differences battery was administered either before or after the eye-movement experiment, individually or in small groups.

Spelling dictation test

This test consists of 20 words selected from a larger set administered to samples of Australian university students by Burt and Tate ( 2002 ) to cover a broad range of spelling accuracy. In their sample, the words were correctly spelled by between 35% and 92% of participants, and discriminated between above-average and below-average spellers. The experimenter read aloud each word and included it in a short sentence to resolve any ambiguity. Most participants handwrote the word on a response sheet, but one subsample ( n  = 62) completed a computerized version of the test (implemented in Qualtrics but administered in the laboratory) in which they typed the word into a response box. Handwritten responses were scored manually by the experimenter, and typed responses were automatically scored by Qualtrics. Administration of the spelling dictation test typically took between 5 and 7 minutes.

Spelling recognition test

This test consists of 88 items, half correctly spelled and half incorrect. Incorrect spellings were constructed to be phonologically plausible to increase the difficulty of the test and encourage reliance on orthographic knowledge (Katz & Frost, 2001 ). Participants were given unlimited time to select all incorrectly spelled items. They viewed all the items together in four columns and recorded their responses either by circling items on paper or by selecting items in a Qualtrics survey. Responses were manually scored by the experimenter or automatically scored in Qualtrics. Scores on the spelling recognition test are calculated out of a total possible score of 88, subtracting the number of correct spellings selected (i.e., false alarms) and the number of incorrectly spelled items that were not selected (i.e., misses). Participants typically took between 5 and 10 minutes to complete the spelling recognition test.

Nelson-Denny Test (NDT)

All participants completed Form H of this instrument, which includes two subtests. In the 80-item vocabulary test, participants are given a word and asked to select the best-matching word or phrase from five options. The separately timed comprehension subtest includes 38 items relating to seven short passages on a range of topics. It also provides an assessment of reading rate by instructing participants to mark their progress through the first passage after 1 minute has elapsed. A total of 361 of the participants were administered the vocabulary and comprehension sections with standard timing procedures (i.e., 15 minutes for vocabulary and 20 minutes for comprehension); the remaining 345 participants were allowed only half the usual time limit for each subtest. To allow a direct comparison of scores under the two administration procedures, a further sample of 107 participants (referred to as the Full+Half sample) completed the full-timed version of the NDT but marked where they were up to at the halfway point for each section so that both full- and half-timed scores could be computed. This sample also completed the two additional tests described below.

Author Recognition Test (ART)

The Full+Half sample of 107 participants completed this test, which required them to identify which of a list of 100 names they recognized to be authors. The test items were taken from Moore and Gordon ( 2015 ), who removed 15 poorly discriminating author names from the ART scale developed by Acheson et al. ( 2008 ), leaving 50 author names and 50 foils. All names were listed in alphabetical order by surname on a response sheet, and participants circled the names they recognized as authors. Participants were instructed not to guess, because they would be penalized for incorrect responses. Administration of the ART took between 3 and 5 minutes. Scores were computed by subtracting the number of false alarms to foil names from the number of correctly selected authors, the standard method of scoring the ART (Acheson et al., 2008 ). Moore and Gordon ( 2015 ) found that this scoring method yielded slightly higher correlations with measures of reading behavior than a measure of hit rate alone.

Phonemic decoding

The Full+Half subsample also completed the phonemic decoding subtest of the TOWRE (Torgesen et al., 1999 ), which consists of a list of 63 nonwords ranging from one to four syllables. After completing a practice list of eight items, participants were given 45 seconds to read aloud as many nonwords as they could. Responses were recorded by a microphone, and the number of correctly pronounced items was checked offline by a research assistant. Pronunciations were deemed correct if they applied plausible grapheme-phoneme correspondences for the complete nonword string. Partial scores were not applied.

Results and discussion

Descriptive statistics.

Table 1 presents the descriptive statistics for each measure – spelling dictation, spelling recognition, vocabulary, reading comprehension, reading rate, author recognition, and phonemic decoding efficiency – separately for the three subsamples administered full-timed and half-timed versions of the NDT.

As expected for a selected sample of predominantly native English-speaking university students, the average level of reading performance was relatively high: the means for the NDT vocabulary and comprehension tests administered with standard timing corresponded to the 74th and 75th percentiles, respectively, of the norms for US students in the first year of a 4-year college program, while the average reading rate corresponded to approximately the 60th percentile of that cohort. The mean and standard deviation of the ART scores were very similar to those obtained by Moore and Gordon ( 2015 ) for the 50-item test (M = 13.75; SD = 6.81), while the average phonemic decoding score was above the 90th percentile for 6th grade children – the oldest members of the Australian normative sample (Marinus, Kohnen, & McArthur, 2013 ). Despite the high average performance, the wide score ranges indicate that the samples showed considerable variability on all measures in the battery.

Unsurprisingly, the NDT vocabulary and comprehension scores were lower for the half-timed version, but vocabulary scores were relatively less affected by the reduced time limit than comprehension. The reasons that the two subtests differ in sensitivity to speed pressure and the implications for their discriminative power are explored below. The manipulation of administration time did not affect the NDT reading rate measure, because it is based on only the first passage of the comprehension test. The similarity between the mean reading speeds of the three subsamples provides reassurance that they are of relatively equivalent reading proficiency.

Internal consistency of spelling tests

To address our first aim of assessing the psychometric properties of the two novel tests of spelling ability, analyses of the internal consistency of each test were conducted.

Spelling dictation

Factor analysis of the 20 items from the spelling dictation test using maximum likelihood estimation to test for unidimensionality showed a clear break in the scree plot after the first factor (from an eigenvalue of 4.56 to 1.21), which accounted for 18.88% of the variance. Footnote 2 Factor loadings ( λ ) and extracted communalities ( h 2 ), which represent the proportion of the item’s variance accounted for by this factor, are shown in Table A1 with descriptive statistics for each item (see Appendix for full test and norms). All but three items ( warranty , asymmetry , diligent ) had factor loadings greater than 0.30, indicating adequate prediction of the item from the latent construct assessed by the test as a whole. Internal consistency indexed by Cronbach’s alpha of 0.814 is in the range classified as ‘good’, and the mean inter-item correlation of 0.18 falls at the lower end of the range taken to indicate test homogeneity (Clark & Watson, 1995 ). Further evidence of homogeneity is provided by the point biserial correlations ( r pb ; see Table A1 ), which index whether the item discriminates between individuals in the same way as the total test. All items had positive point biserial correlation coefficients > 0.30 (65% above 0.40), indicating that the items consistently predicted the total test score.

To provide further insight into the psychometric properties of the test, Rasch analysis of the 20 items was conducted using RUMM2030, yielding the Rasch item difficulty indices (in logits) shown in Table A1 for each item, and summarized in the person-item map in Fig. 1 . Rasch item difficulty indices estimate the level of person ability at which an item has a 50% chance of being correctly/incorrectly endorsed. Higher Rasch difficulty values indicate more difficult items (i.e., a greater level of ability required). The most difficult item ( conciliatory ), which yielded a Rasch index of 2.45, was correctly spelled by less than 20% of the sample, while over 88% correctly spelled the easiest item ( euphoric , Rasch index of −1.94). Rasch analysis also allows items and individuals to be measured on the common scale of Rasch logit units depicted in the person-item map (Fig. 1 ), which plots the participants by ability against the test items by difficulty. In general, the distribution of item difficulty aligns with the distribution of person ability; however, the range of the ability distribution is wider than that of the item difficulty distribution. This indicates that the test items may not effectively capture individual differences in spelling ability at the lowest and highest levels. Footnote 3

Plot of participants (’PERSONS’) by performance on the Spelling Dictation Test against the 20 test items (’ITEMS’) by Rasch difficulty (in logit units). Better test performance and more difficult items appear at the top of the figure

Spelling recognition

Four of the 88 items ( appreciate , distinguish , exhibition , annual ) of the spelling recognition test were identified as correctly spelled by all participants and so were excluded from item-level analyses. The remaining 84 items were factor analyzed using maximum likelihood estimation to check for unidimensionality. Examination of the scree plot showed a clear break after the first factor, which accounted for 8.83% of the variance. Factor loadings ( λ ) and extracted communalities ( h 2 ) are shown in Table A3 , with descriptive statistics for each item (see Appendix for complete test and norms). Overall internal consistency was ‘good’ (Cronbach’s alpha = 0.859). However, the factor loadings for some items were marginally negative (e.g., sufficient , elementary , inhibition ), and many items had low communalities (e.g., attitude , consequence , parallel ), indicating inadequate prediction of the item from the latent construct of spelling recognition ability. The mean inter-item correlation was 0.06, indicating that the test is likely not homogeneous. The point biserial correlations ( r pb ) for many items were also negative or near-zero (e.g. senior , fulcrum , guitar ), indicating poor discrimination of spelling ability. These limitations arise, at least in part, because the test contains too many easy items. Although the percentage of participants correctly identifying each item as either a correct or incorrect spelling ranged from 26.6% to 100%, more than half of the items were correctly classified by over 90% of participants, limiting its sensitivity.

This problem is evident in the person-item map for the spelling recognition test presented in Fig. 2 . The Rasch item difficulty indices (in logits) reveal considerable variability among items: consequence was the easiest, with a Rasch score of −3.55, and vigilant , at 3.71, was the most difficult. Participants were also relatively normally distributed on the Rasch difficulty scale, but the distributions of ability and item difficulty were not aligned. Many items fell below the lowest level of spelling ability and therefore made little contribution to discrimination between people. It is noteworthy that the majority of the items in the easier half of the difficulty range are correct spellings (e.g., 37 of the 41 items with Rasch values below 0), suggesting a bias to classify items as correctly spelled, i.e., participants were much more likely to fail to identify an incorrect spelling than to falsely classify a correct spelling as incorrect.

Plot of participants (’PERSONS’) by test performance on the Spelling Recognition Test against 84 test items (’ITEMS’) by Rasch difficulty (in logit units). Better test performance and more difficult items appear at the top of the figure. Items in italicized font are incorrect spellings

Thus, the results of item-level analyses for the spelling recognition test clearly indicate that there is room for psychometric improvement. This might be achieved by removing items with low communalities, negative factor loadings, and/or negative and near-zero point biserial correlations, but, as elaborated in the Discussion, attention will need to be paid to the distribution of correct and incorrect spelling and response requirements of the recognition task. The present set of items may also be more effective in discriminating between individuals in samples with lower levels of written language proficiency than those included in our studies. Even though a large proportion of the items yielded very high average accuracy, the Rasch item distribution for the easy items is graded, and only four items were correctly classified by all participants. Importantly, even in our restricted sample, scores on the recognition test were highly correlated with spelling dictation, and the factor analyses reported below show that they tapped the same dimension of variability between people.

Effects of administration time on vocabulary and comprehension performance

To investigate our second research question of how differences in administration time influence the estimates of vocabulary and reading comprehension obtained from the NDT, scores were collated separately for participants using the standard ‘full-timed’ procedures ( n  = 361), those tested using ‘half-timed’ procedures ( n  = 345), and the Full+Half NDT sample ( n  = 107) for whom both full-timed and half-timed measures of vocabulary and reading comprehension were obtained. As illustrated in Fig. 3 , the distribution of scores obtained in both tests was highly negatively skewed under the standard full-timed conditions (vocabulary: skew = −0.97; comprehension: skew = −1.54), reducing discrimination at the upper end of the score distribution, but substantially more normal with the half-timed limit conditions (vocabulary: skew = −0.11; comprehension: skew = 0.46). The half-timed procedure therefore increased discrimination among more proficient readers. Similar differences in the distribution of full-timed and half-timed scores were evident in the Full+Half NDT sample. To determine whether the differential discrimination of the half-timed procedure influenced the relationship of vocabulary and comprehension to other measures, the analyses of convergent and divergent validity reported below were conducted separately on the full-timed and half-timed samples.

Frequency distributions of scores on the Vocabulary (upper panels) and Comprehension (lower panels) subtests of the Nelson-Denny Reading Test for participants tested under full-timed versus half-timed administration conditions

Correlations

Table 2 shows the correlations among spelling and reading measures for the samples that completed the NDT under full- and half-timed administration. All measures were moderately to strongly positively correlated, although the relationship of reading rate with the other measures tended to be weaker than those between spelling, vocabulary, and comprehension. The two spelling tests were very highly correlated in both the full-timed and half-timed NDT subsamples, and they showed almost identical strong correlations between vocabulary and comprehension, and similar moderate relationships between these measures and the two spelling tests. The most substantial difference between the samples was a significantly higher correlation between NDT comprehension and reading rate in the half-timed ( r  = 0.51) than the full-timed sample ( r  = 0.30), z  = 3.018, p  = 0.003, suggesting that the relationship between reading speed and comprehension is stronger under time constraints. However, in general, the simple correlations indicate that the half-timed version of the NDT captures similar dimensions of individual differences among skilled readers as the standard full-timed version.

Further support for this conclusion derives from the similar patterns of correlations observed for the full- and half-timed measures in the Full+Half subsample (see Table 3 ). The data for this sample also showed that the additional ART and phonemic decoding scores were only weakly related to each other, but both were at least moderately correlated with all of the measures of spelling and reading proficiency.

Principal component analysis

To evaluate whether these positively correlated measures of linguistic processing can be reduced to a smaller number of independent dimensions, principal component analyses with promax rotation were conducted. Footnote 4 Each analysis was conducted for both the full-timed and half-timed NDT subsamples, excluding the Full+Half subsample to determine whether administration time changed the dimensional structure. Two components were extracted in each analysis.

The first set of analyses was conducted on the spelling dictation, spelling recognition, vocabulary, and comprehension scores that have been used in our previous published investigations of individual differences in masked priming and sentence reading. As summarized in Table 4 , for both full-timed and half-timed subsamples, the two moderately correlated components accounted for around 85% of variance, with the spelling tests forming Component 1, and NDT vocabulary and comprehension forming Component 2. A second set of analyses added reading rate (rightmost columns of Table 4 ) to evaluate whether it moderated the impact of administration time. Again, both full-timed and half-timed subsamples yielded two similar, moderately correlated components on which the two spelling tests formed Component 1. In the full-timed subsample, vocabulary, comprehension, and reading rate again formed Component 2. However, for the half-timed subsample, comprehension and reading rate loaded selectively on Component 2 but vocabulary loaded equally on both components. There was no evidence of a similar cross-component contribution of vocabulary when reading rate was not included as a predictor, suggesting that reading rate captures variance shared with vocabulary that particularly affects performance in half-timed conditions.

A third set of principal component analyses were conducted on the Full+Half subsample. Separate analyses were conducted using full-timed and half-timed measures of vocabulary and comprehension to evaluate the similarities and differences in the structure observed for samples administered with the independent full- and half-timed measures. These analyses also included ART and phonemic decoding scores to confirm and refine understanding of the dimensions of individual differences. Two components were extracted using a promax rotation. Component loadings are shown in Table 5 .

Paralleling the results for the independent samples, analyses including both the full- and half-timed scores yielded moderately correlated components that separated the spelling tests from the other measures: the two spelling tests fell on Component 2, and the NDT measures of vocabulary, comprehension, and reading rate on Component 1. Footnote 5 The two additional individual-differences measures collected for this sample provided useful evidence of convergent and divergent validity because they loaded on different components. The ART index of print exposure loaded on Component 1 with the three NDT measures, while the phonemic decoding score loaded on Component 2 with the two spelling tests. As well as providing converging evidence for the two components of reading proficiency identified in the principal component analyses, the differential loadings of these two measures also provide useful additional information about the nature of the dimensions they assess. The relationship between phonemic decoding and spelling ability is consistent with the view that orthographic precision depends on the amalgamation of orthography and phonology (Ehri, 2015 ; Perfetti, 2007 ), and that this dimension is partially independent of the higher-order linguistic knowledge and skills captured by tests of vocabulary, comprehension, and reading speed, and the broad reading experience tapped by the ART.

General discussion

The central aims of the present research were to assess the internal consistency and validity of two recently developed tests of spelling ability that have been used in several investigations of individual differences among skilled readers and to evaluate whether they assessed a dimension of variability that is not effectively captured by other widely used tests of written language proficiency.

The results confirmed previous evidence that the tests of spelling dictation and spelling recognition were highly correlated across the entire sample of close to 800 skilled readers, and in the three independent sub-samples of 107–361 participants ( r  = 0.75–0.82) defined by the different administration timing conditions of the NDT. There is therefore little evidence that dictation and recognition tests tap independent input and output representations. Rather, spelling appears to be a relatively unitary ability in skilled adult readers, but, as in other domains, production tests like dictation are more difficult than recognition tests. However, the failure to demonstrate independent dimensions of spelling ability may reflect psychometric limitations of the test of spelling recognition discussed below.

Factor analyses showed that the items in each test had good internal consistency (Cronbach’s alpha > 0.8) and assessed a relatively homogeneous dimension of individual variance. Rasch analysis revealed that the item and ability distributions for the spelling dictation test were relatively well aligned, although the items failed to tap either the upper or lower extreme of the ability distribution. The spelling recognition test was more poorly calibrated. The distributions of ability scores for both participants and items were relatively normal, but the test contained too many easy items. The analyses also revealed a strong bias to withhold ‘incorrect’ responses – yielding high accuracy for correct spellings at the expense of a high miss rate for incorrect spellings. This conservative strategy may reflect the response requirement of the present recognition task, i.e., to check incorrect spellings. The opposite instruction – to endorse correct spellings – may yield the reverse bias: participants may be cautious about endorsing spellings as correct and therefore show a high miss rate to avoid making false alarms to incorrect spellings. Such biases might be reduced by telling participants that 50% of the items are correctly spelled. However, a more effective strategy for developing a psychometrically sound spelling recognition test may be to use a forced-choice format in which two alternative spellings of a single item are simultaneously presented to avoid the inherent problems of yes/no procedures.

Both spelling tests could undoubtedly be psychometrically improved to yield better ability-item alignment by adding more difficult items, particularly to the recognition test. However, care would be required to ensure that increasing discrimination did not reduce the specificity of the tests. Difficulty in producing or recognizing a correctly spelled word may reflect low vocabulary or reading experience rather than a specific limitation in the precision of orthographic knowledge for known words. Consistent with this possibility, there were small, significant correlations between Rasch difficulty and log frequency estimates from both the SUBTLEX and HAL corpora for both correctly ( r  = –.24 and −.30, respectively) and incorrectly spelled words ( r  = –.40; –0.37). Adding further items that are sufficiently difficult to be discriminating may therefore risk confounding spelling ability with vocabulary. As discussed in more detail in the next section, the importance of distinguishing the effects of spelling and vocabulary depends on the research goals.

The relatively poor discriminative power of the spelling recognition test may also, in part, reflect the relatively elite population of participants our samples were drawn from: monolingual tertiary students from a metropolitan, research-intensive, high-entry university. The test may yield more effective discrimination between individuals in more diverse samples of adult readers. Nevertheless, even within our restricted samples, it demonstrated strong convergent validity with the spelling dictation test. The two measures were highly correlated, and they defined a separate principal component in the three independent samples of participants, regardless of differences in the administration time of the NDT vocabulary and comprehension tests. Thus, despite its poorer discrimination, specific variance in performance on the recognition test was clearly shared with the dictation test rather than the other measures of written language proficiency.

The inclusion of the additional measures of ART and phonemic decoding in the Full+Half sample contributed to establishing the convergent and divergent validity of the two components. Both measures showed at least moderate simple correlations with the tests defining both factors, despite their low correlation with each other. Confirming that they tapped different dimensions of individual difference, they loaded on different components in analyses using both full- and half-timed NDT measures. Their differential loadings are conceptually consistent with the lexical quality hypothesis. The broad index of reading experience tapped by the ART loaded with other measures of higher-level reading processes, while phonemic decoding ability selectively loaded with the two tests of spelling ability. This evidence that phonemic decoding is specifically related to spelling ability is consistent with the view that knowledge of orthographic-phonological correspondences is a critical foundation for the development of precise orthographic knowledge (Ehri, 2015 ); Nation, 2017 ). The foundational role of the alphabetic principle and phonological decoding in learning to read English has been well established in developmental populations (e.g., Byrne, 1998 ; Share, 1995 ) and has been argued to provide “the means for orthographic learning – the gradual accumulation of orthographic knowledge, via reading experience” (Nation, 2017 , p. 3). The present evidence that phonemic decoding is specifically related to spelling ability even in skilled adult readers provides strong support for such claims.

The consistency of the component structure regardless of whether vocabulary and comprehension were assessed in time-pressured conditions also suggests that the separable dimensions of individual differences exert a stronger influence than the method variance associated with strategic adjustments of speed-accuracy in response to different timing constraints. As summarized in Fig. 3 , reducing the time limits for these two tests had a dramatic influence on the distribution of scores for both subtests of the NDT. Such differences would be expected to influence the discriminative capacity of the tests and compromise their sensitivity to different sources of variability. However, the overall proportion of variance accounted for, and the component structure, was very consistent across variations in administration time. The only observed change in component structure was due to the inclusion of NDT reading rate in the half-timed sample. In this analysis, NDT vocabulary showed similar moderate loadings on both principal components, rather than loading selectively with the higher-level processing measures as it did in all other analyses. No such shift was evident in the analysis using half-timed scores for the Full+Half sample, which was tested under the same longer time limits as the independent full-timed sample but instructed participants to mark where they were up to at the half-timed limit. This sample was therefore not subject to the same time pressure as the half-timed sample. The contribution of vocabulary to the ‘precision’ component therefore appears to be associated with strategies that are enhanced under conditions of speed pressure.

Thus, with respect to the practical goals of this research, the analyses demonstrate that the two tests of spelling ability provide relatively internally consistent, converging measures of a source of individual differences among skilled readers that is partially independent of the more typically used measures of vocabulary, reading comprehension, and speed. Despite some psychometric limitations, these instruments have already yielded new empirical insights, briefly elaborated in the next section, that demonstrate their utility in research on individual differences in reading.

One practical issue that may warrant further investigation concerns the value of assessing both spelling dictation and spelling recognition. The high correlation between the two tests provides little support for the view that spelling production and recognition tap different representations or processes (e.g., Ellis, 1993 ). Nevertheless, the use of two converging measures that involve different encoding and response requirements may contribute to more effectively isolating variance specifically associated with the spelling ability. If a single test was selected, spelling dictation would certainly be favored on psychometric grounds. Assessing production of the complete orthographic form also provides a more conceptually valid index of precise orthographic knowledge and yields errors that can potentially be analyzed to diagnose the source of spelling problems. However, the informal experience we have gained by making the tests available to other research groups is that the spelling recognition test is preferred, presumably because of both its ease of administration and greater acceptability to participants: spelling dictation tests often evoke performance anxiety even in skilled readers. It may be possible to develop a more psychometrically sound spelling recognition test by using a forced-choice rather than yes/no format to reduce the biases of recognition tasks, and selecting distractor items that systematically manipulate different dimensions of similarity to the target (e.g., orthographic, phonological, morphological). Such an approach may allow construction of a recognition test that achieves the same level of discrimination and diagnostic capacity as spelling dictation. The item analyses reported here provide the foundation for such developments.

The role of spelling in assessing lexical quality

The consistent evidence that tests of spelling ability tap a partially independent dimension of variability in reading skill aligns with the precision component of Perfetti’s ( 2007 ) construct of lexical quality. It also confirms Andrews’ ( 2008 , 2012 , 2015 ) proposal that measures of spelling ability can be used to tap lexical precision and account for unique variance that is not captured by the measures of reading comprehension and vocabulary more typically used to assess individual differences among skilled readers.

Lexical quality is defined by both the precision of orthographic representations and the coherent, synchronous activation of their associated phonological and semantic codes (Andrews, 2015 ; Perfetti, 2007 ). The second component consistently identified in the principal component analyses was defined by measures of higher-level knowledge and processes: the semantic knowledge indexed by vocabulary, text-level measures of reading comprehension and speed, and the ART index of reading experience. It therefore appears to tap the processes associated with efficient, synchronous retrieval and use of the complex of lexical codes associated with a specific word form – processes related to the ‘coherence’ dimension of lexical quality (Perfetti, 2007 ). This is consistent with evidence that individuals with higher scores on the coherence than the precision component show stronger semantic influences of masked primes in both lexical decision (Andrews & Lo, 2013 ) and semantic categorization tasks (Andrews et al., 2017 ). The present evidence that NDT vocabulary loaded equally on both the precision and coherence factors for the half-timed sample only when reading rate was included as a predictor suggests that knowledge of word meanings combines with high orthographic precision to facilitate rapid, coherent retrieval of the complex of lexical codes that define a word, as well as contributing to higher-level word integration processes. It also implies that the contribution of the coherence between orthographic and semantic codes to early lexical retrieval may be more effectively tapped when vocabulary is assessed under speed pressure.

The multidimensional nature of lexical quality revealed by these analyses highlights the value of including multiple measures of lexical proficiency in studies designed to investigate individual differences among highly skilled readers. As confirmed by the present data, measures of written language proficiency are generally at least moderately inter-correlated, so a single reliable measure can potentially tap variability in the broad dimension of lexical proficiency. For example, Yap’s systematic investigations of the role of vocabulary in predicting individual differences across a range of tasks including single-word lexical decision (Yap et al., 2012 ) and semantic categorization tasks (Pexman & Yap, 2018 ) and studies of unmasked (Yap et al., 2009 ) and masked priming (Tan & Yap, 2016 ) have been interpreted as demonstrating the role of the “integrity” of readers’ lexical representations in facilitating automatic lexical retrieval. However, vocabulary is also correlated with measures of higher-order linguistic processing such as listening comprehension (Braze et al., 2016 ). Similarly, in the present analyses, vocabulary most consistently loaded with the text-based measures of reading comprehension and reading rate rather than with spelling ability. The consistency of the modulating effects of vocabulary across a range of word-identification tasks may reflect the fact that it contributes to both the precision and coherence components of lexical quality, at least when it is assessed under speeded testing conditions. Vocabulary tests may therefore be a useful index of the broad construct of lexical quality when researchers are limited to a single measure.

However, studies using a broader test battery that allows different facets of lexical quality to be separated provide additional insights into how lexical quality contributes to effective reading that are obscured by more global measures. Investigations of masked orthographic priming of lexical-decision performance have shown that spelling selectively predicts competition from similar words on both behavioral performance (Andrews & Hersch, 2010 ; Andrews & Lo, 2012 ) and the N400 component of the event-related potential waveform (Meade et al., 2018 ). An ‘orthographic profile’ of relatively higher spelling than vocabulary is also associated with form-based morphological decomposition indexed by equivalently strong masked priming for genuine (e.g., hunter-HUNT ) and pseudo-morphemically related ( corner-CORN) prime-target pairs (Andrews & Lo, 2013 ). In contrast, the opposite ‘semantic profile’ of higher vocabulary than spelling ability predicts strong masked priming for genuinely morphological pairs (Andrews & Lo, 2013 ) and stronger semantic congruence priming in a semantic categorization task (Andrews et al., 2017 ).

Isolating the different dimensions of lexical quality has also yielded new insights into the processes underlying skilled readers’ eye movements during sentence reading. Reading comprehension is typically the best predictor of the overall speed and efficiency of sentence reading, consistently predicting shorter fixation durations and sentence/passage reading times, but spelling ability selectively predicts measures related to eye-movement control such as saccade length (Veldre & Andrews, 2014 ) and word skipping (Slattery & Yates, 2018 ; Veldre & Andrews, 2016b ; Veldre et al., 2017 ). The combination of spelling and reading comprehension is also a stronger predictor of the extent and depth of parafoveal processing than either measure alone (Veldre & Andrews, 2015a , 2015b ). However, the two measures have counteracting effects on semantic processing of parafoveal previews: higher reading comprehension was associated with stronger preview benefit from a semantically related or contextually plausible word, whereas better spellers showed reduced semantic/plausibility preview benefit relative to poorer spellers (Veldre & Andrews, 2016a , 2016b ; see Andrews & Veldre, 2019 , for a review). These data converge with the masked morphological and semantic priming results in suggesting that high spelling ability supports rapid retrieval of a word’s orthographic form. In the masked orthographic priming paradigm, this yields competition from word primes that are orthographically similar to the target. This competition eliminates the benefits of sublexical overlap between the prime and target which is responsible for priming in poor spellers. In studies of parafoveal preview, the conflict between the orthographic form of the preview and the target eliminates the benefits of the preview’s contextual acceptability – which is the source of semantic preview benefit in poorer spellers. In both cases, the detrimental effects of high orthographic precision are due to the conflicting perceptual input introduced by the masked prime, or parafoveal preview. In normal reading, where such misleading information is not presented, rapid retrieval of either orthographic or semantic features of a briefly presented, or parafoveal, word will facilitate lexical retrieval and enhance the efficiency of reading.

This evidence of systematic individual differences among skilled readers reviewed above contributes to resolving some of the contradictory findings obtained in typical analyses of the averaged data for samples of skilled readers that have sustained ongoing debates about visual word recognition and reading (Andrews, 2012 ). The evidence that spelling ability selectively modulates lexical competition suggests that efforts to ‘crack the orthographic code’ (emphasis added) for reading (e.g., Grainger, 2008 ) may be misguided. Similarly, the differential effects of vocabulary and spelling on morphological and semantic priming suggest that there may not be a single answer to the debate between form-first versus cascaded activation accounts of semantic retrieval (e.g., Forster, 2013 ; Rodd, 2004 ). Individual differences also influence the interactions between foveal and parafoveal processing that provide a critical source of evidence for distinguishing between serial and parallel models of eye movement control (e.g., Engbert, Nuthmann, Richter, & Kliegl, 2005 ; Reichle, Pollatsek, Fisher, & Rayner, 1998 ; Snell, van Liepsig, Grainger, & Meeter, 2018 ).

It is important to emphasize that the different components of lexical quality are not entirely independent. Across each of our samples, the correlation between the two rotated components identified in our factor analyses was consistently around 0.5. In unrotated factor solutions, all the NDT and spelling tests loaded on the first common factor, which accounted for 54% and 56% of variance in the full- and half-timed samples, respectively; while the discrepancies between spelling and NDT measures emerged on the second component, which captured an additional 19–20% of variance. Thus, the primary dimension of individual differences tapped by our test battery is common to spelling, vocabulary, and reading comprehension – an index we have referred to as ‘overall proficiency’ (e.g., Andrews, 2015 ; Andrews & Veldre, 2019 ). However, when this shared variance is partialed out, a second, weaker dimension of variance is revealed that is defined by discrepancies between individuals’ level of spelling and vocabulary/comprehension ability.

The fact that spelling taps a common dimension of variance that is shared with vocabulary and comprehension is not surprising. Knowledge of the meanings and the spellings of words both depend critically on experience reading (and writing). Although spoken language experience clearly makes an important contribution to vocabulary acquisition, the majority of children’s vocabulary learning occurs through reading (Nagy, Herman, & Anderson, 1985 ), and print exposure is clearly essential for spelling in both children and adults (e.g., Mol & Bus, 2011 ) – particularly in the quasi-regular writing system of English. Footnote 6 Nation ( 2017 ) argues that repeated exposure to words across a diversity of contexts, episodes, and experiences supports the development of a rich interconnected database and leads to “local variation at the word level: a ‘lexical legacy’ that is measurable during word reading behavior” (p. 1). Vocabulary and spelling knowledge both depend on the capacity to extract invariant patterns from text and establish connections both between the different components of a lexical form (orthography, phonology, and semantics) and between related linguistic units. Individual differences in these capacities, and their consequences for the quality of the lexical database that is acquired through reading experience, may account for the shared variance in spelling, vocabulary, and reading. However, the evidence of a second dimension of individual variability that is tapped by the discrepancy between orthographic and semantic knowledge implicates an additional source of variation in either the knowledge extracted from reading experience, or the manner in which it is applied to word identification and reading. Further systematic investigations of orthographic learning in developing readers (e.g., Nation, 2017 ; Joesph, Wonnacot, Forbes, & Nation, 2014 ; Tamura, Castles, & Nation, 2017 ) and computational modeling of this development (e.g., Ziegler, Perry, & Zorzi, 2014 ) will contribute to determining the individual, instructional, and contextual factors responsible for these differences.

Conclusions

The spelling dictation test and spelling recognition test described here each take only approximately 5 minutes to administer but capture unique variance among the population of skilled, college-aged readers that has been shown to modulate effects in masked priming and sentence reading tasks. Both spelling tests were unidimensional, displayed good internal consistency, correlated with other measures of reading ability, and formed a distinct component across a large college-aged sample. The item-level analyses reported in the present paper provide a basis for further refinement of these instruments, particularly the spelling recognition test, to better discriminate among participants at the highest levels of proficiency, but the present versions may discriminate more effectively in less-skilled adult samples. The results also demonstrate that using shorter time limits for standard tests of vocabulary and comprehension enhances discrimination among skilled readers without substantially changing the relationships between different components of written language proficiency. These practical contributions to measuring lexical quality will support the development of a richer body of empirical evidence about how individual differences modulate skilled word recognition and reading. Incorporating such variation into models of these processes is widely acknowledged to be a critical step for future theoretical development in understanding these essential educational and vocational skills (e.g., Andrews & Reichle, 2019 ; Radach & Kennedy, 2013 ; Rayner, Abbott, & Plummer, 2015 ).

Demographic data were not collected from all participants. The summary statistics reported here are based on between 466 and 609 participants, but are representative of all samples tested.

Confirming that a single-factor solution captured the majority of predictable item variance, a second factor analysis that extracted two factors revealed that they accounted for 19.04% and 2.42% of variance, respectively, and were highly correlated (r = 0.678).

To evaluate whether these limitations in discrimination were due to categorical scoring of spellings as correct/incorrect, the dictation responses for a subset of 125 participants were rescored by calculating the Levenshtein distance between the correct spelling and the participant’s response using the vwr package (Keuleers, 2013 ) in R (R Core Team, 2019 ). On average, participants’ spellings were very similar to the correct spelling – the mean edit distance was less than one letter (M = 0.66, SD = 0.46). Levenshtein-based scores of spelling accuracy were also highly correlated with the categorical dictation scores by both subjects ( r  = −0.78) and items ( r  = −0.95), and with the Rasch difficulty index based on categorical scores ( r  = −0.97). The Levenshtein scores also showed similar, albeit slightly weaker, relationships as the categorical scores, with variability in NDT vocabulary (−0.42 vs. 0.51), comprehension (−0.36 vs. 0.46), and reading rate (−0.29 vs. 0.37), and spelling recognition (−0.58 vs. 0.80). Thus, at least for this set of items, the greater precision offered by Levenshtein distance measures does not appear to improve discrimination among skilled readers, presumably because their spelling errors are generally limited to specific ambiguous phoneme-grapheme correspondences.

All principal component analyses were also conducted with no rotation. Component structure prior to rotation also yielded two components in which the first centroid had high loadings from all variables, and the second differentiated the spelling tests from the other measures.

The same two-factor structure was observed in analyses including the smaller subset of predictors available for the full- and half-timed samples (see Appendix Tables A5 and A6 ). NDT vocabulary loaded selectively on the same component as NDT comprehension both when NDT reading rate was and was not included as a predictor.

An issue that is beyond the scope of this paper, but warrants further research, is whether the independent effects of spelling ability on reading behavior generalize to languages other than English. The construction of precise word-specific orthographic representations may be a specific response to the notoriously inconsistent and idiosyncratic spelling-sound correspondences of English which have been demonstrated to be more complex than any other alphabetic orthography (Share, 2008 ). This complexity may drive extraction of multi-letter units at a range of different ‘grain-sizes’ (Ziegler & Goswami, 2005) to capture systematic, higher-order consistencies in the mapping between orthography and phonology (e.g., Kessler & Treiman, 2001 ). If so, the independent contributions of spelling ability to predicting reading behavior may be limited to ‘orthographically deep’ scripts (Frost, Katz, & Bentin, 1987 ) but absent in transparent alphabetic script like Finnish and Spanish, for which word-specific orthographic representations may be unnecessary or redundant. Alternatively, the contribution of orthographic knowledge to reading behavior may reflect general cognitive principles of skill acquisition (Anderson, 1981 ). Across a range of domains of skilled human behavior, the transition from novice to expert performance is characterized by a shift from slow, deliberate, effortful, algorithmic processing to rapid, automatic performance mediated by direct-retrieval mechanisms (e.g., Logan, 1988 ). From this perspective, unitized orthographic representations may be a signature of expert word recognition and reading even in highly transparent scripts. However, spelling tests like those reported here are unlikely to provide a sensitive measure of orthographic knowledge in such languages, so other methods will be required in order to investigate this possibility.

Acheson, D.J., Wells, J.B. & MacDonald, M.C. (2008). New and updated tests of print exposure and reading abilities in college students. Behavior Research Methods, 40, 278–289.

PubMed Central   Google Scholar  

Adelman, J. S., Johnson, R. L., McCormick, S. F., McKague, M., Kinoshita, S., Bowers, J. S., . . . Davis, C. J. (2014). A behavioral database for masked form priming. Behavior Research Methods, 46 , 1052–1067.

Anderson, J.R. (1981). Cognitive skills and their acquisition. Hillsdale, NJ: Erlbaum.

Google Scholar  

Andrews, S. (2008). Lexical expertise and reading skill. In B. H. Ross (Ed.), The Psychology of Learning and Motivation: Advances in Research and Theory (Vol. 49, pp. 247–281). San Diego, CA: Elsevier.

Andrews, S. (2012). Individual differences in skilled visual word recognition and reading: The role of lexical quality. In J. S. Adelman (Ed.), Visual Word Recognition (Vol. 2, pp. 151–172). London: Psychology Press.

Andrews, S. (2015). Individual differences among skilled readers: The role of lexical quality. In A. Pollatsek & R. Treiman (Eds.), The Oxford Handbook of Reading (pp. 129–148): Oxford University Press.

Andrews, S. & Hersch, J. (2010). Lexical precision in skilled readers: Individual differences in masked neighbor priming. Journal of Experimental Psychology: General, 139 , 299–318.

Andrews, S. & Lo, S. (2012). Not all skilled readers have cracked the code: Individual differences in masked form priming. Journal of Experimental Psychology: Learning, Memory, and Cognition, 38 , 152–163.

PubMed   Google Scholar  

Andrews, S. & Lo, S. (2013). Is morphological priming stronger for transparent than opaque words? It depends on individual differences in spelling and vocabulary. Journal of Memory and Language, 68 , 279–296.

Andrews, S., Lo, S., & Xia, V. (2017). Individual differences in automatic semantic priming. Journal of Experimental Psychology: Human Perception and Performance, 43 , 1025–1039.

Andrews, S., & Reichle, E.D. (2019). The cognitive architecture of reading: The organization of an acquired skill. In P. Hagoort (Ed)., Human Language: From genes and brains to behaviour . MIT Press.

Andrews, S., & Veldre, A. (2019). What is the most plausible account of the role of parafoveal processing in reading? Language and Linguistics Compass .

Ashby, J., Rayner, K., & Clifton, C. (2005). Eye movements of highly skilled and average readers: Differential effects of frequency and predictability. Quarterly Journal of Experimental Psychology, 58 , 1065–1086.

Braze, D., Katz, L., Magnuson, J.S., Mencl, W.E., Tabor, W., Van Dyke, J.A. et al., (2016). Vocabulary does not complicate the simple view of reading. Reading and Writing, 29 , 435–451.

Braze, D., Tabor, W., Shankweiler, D., & Mencl, W.E. (2007). Speaking up for vocabulary: Reading skill differences in young adults. Journal of Learning Disabilities, 40, 226–243.

PubMed   PubMed Central   Google Scholar  

Breadmore, H.L., & Deacon, S.H. (2019). Morphological processing before and during children’s spelling. Scientific Studies of Reading, 23, 178–191.

Brown, J. I., Fishco, V. V., & Hanna, G. (1993). Nelson-Denny Reading Test . Rolling Meadows, IL: Riverside Publishing Company.

Burt, J. S., & Tate, H. (2002). Does a reading lexicon provide orthographic representations for spelling? Journal of Memory and Language, 46 , 518–543.

Byrne, B. (1998). The foundation of literacy: the child’s acquisition of the alphabetic principle . Hove, UK: Psychology Press.

Byrne, B., Coventry, W. L., Olson, R. K., Hulslander, J., Wadsworth, S., DeFries, J. C., . . . Samuelsson, S. (2008). A behaviour-genetic analysis of orthographic learning, spelling and decoding. Journal of Research in Reading, 31 , 8–21.

Clark, L. A., & Watson, D. (1995). Constructing validity: Basic issues in objective scale development. Psychological Assessment, 7 , 309–319.

R Core Team. (2019). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org/

Dehaene, S. (2009). Reading in the brain . New York: Penguin Group.

Dehaene, S., Pegado, F., Braga, L., Ventura, P., Filho. G., Jobert, A., Dehaene-Lambertz, G., Kolinsky, R., Morais, M. & Cohen, L. (2010). How learning to read changes the cortical networks for vision and language. Science, 3, 1359–1364.

Drieghe, D., Veldre, A., Fitzsimmons, G., Ashby, J., & Andrews, S. (2019). The influence of number of syllables on word skipping during reading revisited. Psychonomic Bulletin & Review, 26 , 616–621.

Ehri, L. (2000). Learning to read and learning to spell: Two sides of a coin. Topics in Language Disorders, 20, 19–37.

Ehri, L. (2015). How children learn to read words. In A. Pollatsek & R. Treiman (Eds.), The Oxford handbook of reading (pp. 293–310), Oxford, England: Oxford University Press

Ellis, A. W. (1993). Reading, writing and dyslexia: A cognitive analysis . Hillsdale, NJ: Lawrence Erlbaum Associates.

Emmorey, K., Midgley, K.J., Kohen, C.B., Sevcikova Sehyr, Z. & Holcomb, P.J. (2017). The N170 ERP component differs in laterality, distribution and association with continuous reading measures for deaf and hearing readers. Neuropsychologia, 106, 298–309.

Engbert, R., Nuthmann, A., Richter, E., & Kliegl, R. (2005). SWIFT: A dynamical model of saccade generation during reading. Psychological Review , 112 , 777–813.

Eskenazi, M. A., Swischuk, N. K., Folk, J. R., & Abraham, A. N. (2018). Uninformative contexts support word learning for high-skill spellers. Journal of Experimental Psychology: Learning, Memory, and Cognition, 44 (12), 2019–2025.

Falkauskas, K. & Kuperman, V. (2015). When experience meets language statistics: Individual variability in processing English compound words. Journal of Experimental Psychology. Learning, Memory, and Cognition, 41, 1607–1627.

Forster, K.I. (2013). How many words can we read at once? More intervenor effects in masked priming. Journal of Memory and Language, 69, 563–573.

Frith, U. (1980). Unexpected spelling problems. In U. Frith (Ed.), Cognitive Processes in Spelling (pp. 495–515). London: Academic Press.

Frith, U. (1986). A developmental framework for developmental dyslexia. Annals of Dyslexia, 36 , 69–81.

Frost, R., Katz, L., & Bentin, S. (1987). Strategies for visual word recognition and orthographical depth: A multilingual comparison. Journal of Experimental Psychology: Human Perception and Performance , 13 , 104–115.

Grainger, J. (2008). Cracking the orthographic code: An introduction. Language & Cognitive Processes, 23, 1–35.

Holmes, V. M., & Babauta, M. L. (2005). Single or dual representations for reading and spelling? Reading and Writing, 18 , 257–280.

Jackson, M. D., & McClelland, J. L. (1975). Sensory and cognitive determinants of reading speed. Journal of Verbal Learning and Verbal Behavior, 14 , 565–574.

Joesph, H.S., Wonnacot, E., Forbes, P., & Nation, K. (2014). Becoming a written word: Eye movements reveal order of acquisition effects following incidental exposure to new words during silent reading. Cognition, 133, 238–248.

Katz, L. & Frost, S.J. (2001). Phonology constrains the internal orthographic representation. Reading and Writing, 14, 297–332.

Kessler, B., & Treiman, R. (2001). Relationships between sounds and letters in English monosyllables. Journal of Memory and Language, 44, 592–617.

Keuleers, E. (2013). vwr: Useful functions for visual word recognition research (R package version 0.3.0). https://CRAN.R-project.org/package=vwr

Kuperman, V. & Van Dyke, J.A. (2011). Effects of individual differences in verbal skills on eye-movement patterns during sentence reading. Journal of Memory and Language, 65, 42–73.

Landi, N., & Perfetti, C. A. (2007). An electrophysiological investigation of semantic and phonological processing in skilled and less-skilled comprehenders. Brain and Language, 102 , 30–45.

Lefly, D. L., & Pennington, B. F. (2000). Reliability and validity of the Adult Reading History Questionnaire. Journal of Learning Disabilities, 33 , 286–296.

Logan, G.D. (1988). Toward an instance theory of automatization. Psychological Review, 95, 492–527.

Luke, S. G., Henderson, J. M., & Ferreira, F. (2015). Children's eye-movements during reading reflect the quality of lexical representations: An individual differences approach. Journal of Experimental Psychology: Learning, Memory, and Cognition, 41 , 1675–1683.

Marinus, E., Kohnen, S., & McArthur, G. (2013). Australian comparison data for the Test of Word Reading Efficiency (TOWRE). Australian Journal of Learning Difficulties, 18 , 199–212.

Meade, G., Grainger, J., Midgley, K.J., Emmorey, K. & Holcomb, P.J. (2018). From sublexical facilitation to lexical competition: ERP effects of masked neighbor priming. Brain Research, 1685, 29–41.

Mol, S.E. & Bus, A.G. (2011). To read or not to read: a meta-analysis of print exposure from infancy to early adulthood. Psychological Bulletin, 137, 267–296.

Moore, M. & Gordon, P. (2015). Reading ability and print exposure: item response theory analysis of the author recognition test. Behavior Research Methods, 47, 1095–1109.

Nagy, W.E., Herman P.A., & Anderson, R.C. (1985). Learning words from context. Reading Research Quarterly, 20, 439–458.

Nation, K. (2017). Nurturing a lexical legacy: reading experience is critical for the development of reading skill. npj Science of Leaning, 2 :3.

Nouwens, S., Groen, M. A., Kleeman, T., & Verhoeven, L. (2017). The role of semantic retrieval in children's reading comprehension development in the upper primary grades. Journal of Research in Reading, 41 , 597–614.

Perfetti, C. & Hart, L. (2001). The lexical basis of comprehension skill. In D. Gorfein (Ed.), On the consequences of meaning selection: Perspectives on resolving lexical ambiguity (pp. 67–86). Washington DC: American Psychological Association.

Perfetti, C. A. (1985). Reading Ability . New York: Oxford University Press.

Perfetti, C. A. (2007). Reading ability: Lexical quality to comprehension. Scientific Studies of Reading, 11 , 357–383.

Perfetti, C. A., & Hart, L. (2002). The lexical quality hypothesis. In L. Verhoeven, C. Elbro & P. Reitsma (Eds.), Precursors of functional literacy (pp. 189–214). Amsterdam, The Netherlands: John Benjamins Publishing.

Pexman, P.M. & Yap, M.J. (2018). Individual differences in semantic processing: Insights from the Calgary Semantic Decision Project. Journal of Experimental Psychology. Learning, Memory, and Cognition , 44 , 1091–1112.

Protopapas, A., Sideridis, G. G., Mouzaki, A., & Simos, P. G. (2007). Development of lexical mediation in the relation between reading comprehension and word reading skills in Greek. Scientific Studies of Reading, 11 , 165–197.

Radach, R., & Kennedy, A. (2013). Eye movements in reading: Some theoretical context. Quarterly Journal of Experimental Psychology, 66 , 429–452.

Rayner, K., Abbott, M. J., & Plummer, P. (2015). Individual differences in perceptual processing and eye movements in reading. In P. Afflerbach (Ed.), Handbook of individual differences in reading: Reader, text, and context (pp. 348–363). New York, NY: Routledge.

Reichle, E. D., Pollatsek, A., Fisher, D. L., & Rayner, K. (1998). Toward a model of eye movement control in reading. Psychological Review,105, 125–157.

Rodd, J.M. (2004). When do leotards get their spots? Semantic activation of lexical neighbors in visual word recognition. Psychonomic Bulletin & Review, 11, 434–439.

Rossi, M., Martin-Chang, S., & Ouellette, G. (2019). Exploring the space between good and poor spelling: Orthographic quality and reading speed. Scientific Studies of Reading, 23, 192–201.

Share, D.L. (1995). Phonological recoding and self-teaching: Sine qua non of reading acquisition. Cognition, 55, 151–218.

Share, D.L. (2008). On the anglocentricities of current reading research and practice: The perils of overreliance on an “outlier” orthography. Psychological Bulletin, 134 , 584–615.

Slattery, T.J., & Yates, M. (2018). Word skipping: Effects of word length, predictability, spelling and reading skill. Quarterly Journal of Experimental Psychology, 71 , 250–259.

Snell, J., van Liepsig, S., Grainger, J., & Meeter, M. (2018). OB1-reader: A model of word recognition and eye movements in text reading. Psychological Review, 125 , 969–984.

Stanovich, K. E., & West, R. F. (1989). Exposure to print and orthographic processing. Reading Research Quarterly, 24 , 402–433.

Swart, N. M., Muijselaar, M. M. L., Steenbeek-Planting, E. G., Droop, M., de Jong, P. F., & Verhoeven, L. (2017). Differential lexical predictors of reading comprehension in fourth graders. Reading and Writing, 30 , 489–507.

Tamura, N., Castles, A., & Nation, K. (2017). Orthographic learning, fast and slow: Lexical competition effects reveal the time course of word learning in developing readers. Cognition, 163, 93–102.

Tan, L. C., & Yap, M. J. (2016). Are individual differences in masked repetition and semantic priming reliable? Visual Cognition, 24 , 182–200.

Taylor, J. N., & Perfetti, C. A. (2016). Eye movements reveal readers’ lexical quality and reading experience. Reading and Writing, 29 , 1069–1103.

Torgesen, J. K., Wagner, R. K., & Rashotte, C. A. (1999). Test of word reading efficiency . Austin, TX: ProEd.

Treiman, R. (2017). Learning to spell: Phonology and beyond. Cognitive Neuropsychology, 34 , 83–93.

Veldre, A. & Andrews, S. (2014). Lexical quality and eye movements: Individual differences in the perceptual span of skilled adult readers. Quarterly Journal of Experimental Psychology, 67 , 703–727.

Veldre, A. & Andrews, S. (2015a). Parafoveal lexical activation depends on skilled reading proficiency. Journal of Experimental Psychology. Learning, Memory, and Cognition , 41 , 586–595.

Veldre, A. & Andrews, S. (2015b). Parafoveal lexical activation is modulated by the precision of skilled readers’ lexical representations. Journal of Experimental Psychology: Human Perception and Performance , 41, 219–232.

Veldre, A., & Andrews, S. (2016a). Semantic preview benefit in English: Individual differences in the extraction and use of parafoveal semantic information. Journal of Experimental Psychology: Learning, Memory, and Cognition, 42 , 837–854.

Veldre, A., & Andrews, S. (2016b). Is semantic preview benefit due to relatedness or plausibility? Journal of Experimental Psychology: Human Perception and Performance, 42 , 939–952.

Veldre, A., Drieghe, D., & Andrews, S. (2017). Spelling ability selectively predicts the magnitude of disruption in unspaced text reading. Journal of Experimental Psychology: Human Perception and Performance, 43 , 1612–1628.

Wilkinson, G.S., & Robertson, G.J. (2017). Wide Range Achievement Test (5 th edition). Pearson Education Inc.

Yap, M.J., Balota, D.A., Sibley, D.E. & Ratcliff, R. (2012). Individual differences in visual word recognition: Insights from the English Lexicon Project. Journal of Experimental Psychology: Human Perception and Performance, 38 , 53–79.

Yap, M. J., Tse, C. S., & Balota, D. A. (2009). Individual differences in the joint effects of semantic priming and word frequency revealed by RT distributional analyses: The role of lexical integrity. Journal of Memory and Language, 61 , 303–325.

Ziegler, J.C., Perry, C., & Zorzi, M. (2014). Modelling reading development through phonological decoding and self-teaching: implications for dyslexia. Philosophical Transactions of the Royal Society B, 369, 20120397 .

Download references

Author information

Authors and affiliations.

School of Psychology, The University of Sydney, Sydney, NSW, 2006, Australia

Sally Andrews, Aaron Veldre & Indako E. Clarke

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Sally Andrews .

Additional information

The data, test instruments, and norms for the Spelling Dictation Test and the Spelling Recognition Test are available at https://osf.io/t4x7r/ . This research was supported under the Australian Research Council’s Discovery Projects funding scheme (project numbers DP16203224, DP190100719).

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Spelling Dictation Test

Administration instructions: The experimenter reads the word aloud to the participant, followed by the sentence containing the word.

1. ABSTINENCE

The ex-alcoholic found it very difficult to maintain complete abstinence from drinking.

2. ACQUAINTANCE

She knew the woman as an acquaintance, but she was not a close friend.

3. DIGESTIBLE

The nurses had to blend the food to make it digestible for the patient.

4. CONCILIATORY

She tried to adopt a conciliatory approach to avoid further conflict.

5. PISTACHIO

The biscuit with pistachio nuts was delicious.

6. WARRANTY

The TV seemed a good buy because it had a 3 year warranty.

7. RHEUMATIC

The old woman suffered rheumatic pain in all of her joints.

8. CRESCENDO

The music reached a crescendo towards the end of the symphony.

9. ASYMMETRY

She found the asymmetry of the design very appealing.

10. AFFLUENT

By comparison with most Asian countries, Australia is very affluent.

11. DILIGENT

Most of the students are lazy but this boy is very diligent in completing his work.

12. AGGRAVATION

The noise from the next classroom was a constant aggravation to the teacher.

13. COLLOQUIAL

It is usually not appropriate to use colloquial language in an essay.

14. EUPHORIC

The student felt euphoric when she completed her last exam.

15. BROCCOLI

Children often dislike broccoli and other green vegetables.

16. SOMERSAULT

The gymnast turned a somersault before landing back on the bar.

17. OBLIVION

Many sportsmen achieve fame when they are young but then sink into oblivion.

18. RHYTHMICAL

The rhythmical beat of the drums was mesmerizing.

19. RAVENOUS

She had missed lunch so by dinner time she felt ravenous.

20. PERSUADE

She tried to persuade him to her point of view.

Spelling Recognition Test

Administration instructions: The participant is given unlimited time to select all the incorrectly spelled items. Note: The item ‘behaviour’ should be replaced by ‘behavior’ if administering the test to speakers of US English.

PLEASE CIRCLE ALL ITEMS BELOW THAT YOU THINK ARE SPELLED INCORRECTLY

Rights and permissions

Reprints and permissions

About this article

Andrews, S., Veldre, A. & Clarke, I.E. Measuring Lexical Quality: The Role of Spelling Ability. Behav Res 52 , 2257–2282 (2020). https://doi.org/10.3758/s13428-020-01387-3

Download citation

Published : 13 April 2020

Issue Date : December 2020

DOI : https://doi.org/10.3758/s13428-020-01387-3

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Reading ability
• spelling ability
• individual differences
• lexical quality hypothesis
• Find a journal
• Publish with us
• Track your research

Help | Advanced Search

Computer Science > Computation and Language

Title: deep lexical hypothesis: identifying personality structure in natural language.

Abstract: Recent advances in natural language processing (NLP) have produced general models that can perform complex tasks such as summarizing long passages and translating across languages. Here, we introduce a method to extract adjective similarities from language models as done with survey-based ratings in traditional psycholexical studies but using millions of times more text in a natural setting. The correlational structure produced through this method is highly similar to that of self- and other-ratings of 435 terms reported by Saucier and Goldberg (1996a). The first three unrotated factors produced using NLP are congruent with those in survey data, with coefficients of 0.89, 0.79, and 0.79. This structure is robust to many modeling decisions: adjective set, including those with 1,710 terms (Goldberg, 1982) and 18,000 terms (Allport & Odbert, 1936); the query used to extract correlations; and language model. Notably, Neuroticism and Openness are only weakly and inconsistently recovered. This is a new source of signal that is closer to the original (semantic) vision of the Lexical Hypothesis. The method can be applied where surveys cannot: in dozens of languages simultaneously, with tens of thousands of items, on historical text, and at extremely large scale for little cost. The code is made public to facilitate reproduction and fast iteration in new directions of research.

Submission history

Access paper:.

• Other Formats

References & Citations

• Google Scholar
• Semantic Scholar

BibTeX formatted citation

Bibliographic and Citation Tools

Code, data and media associated with this article, recommenders and search tools.

• Institution

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs .

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• HHS Author Manuscripts

Lexical knowledge without a lexicon?

Jeffrey l. elman.

University of California, San Diego, La Jolla, CA

Although for many years a sharp distinction has been made in language research between rules and words — with primary interest on rules — this distinction is now blurred in many theories. If anything, the focus of attention has shifted in recent years in favor of words. Results from many different areas of language research suggest that the lexicon is representationally rich, that it is the source of much productive behavior, and that lexically specific information plays a critical and early role in the interpretation of grammatical structure. But how much information can or should be placed in the lexicon? This is the question I address here. I review a set of studies whose results indicate that event knowledge plays a significant role in early stages of sentence processing and structural analysis. This poses a conundrum for traditional views of the lexicon. Either the lexicon must be expanded to include factors that do not plausibly seem to belong there; or else virtually all information about word meaning is removed, leaving the lexicon impoverished. I suggest a third alternative, which provides a way to account for lexical knowledge without a mental lexicon.

Introduction

Words have had a checkered past, at least as objects for scientific study. For many decades, the study of words — their history, meaning, usage, etc. — constituted a main focus of linguistic research. This changed radically in the middle of the last century, starting with the publication of Chomsky’s Syntactic Structures and Aspects of a Theory of Syntax (1957 , 1965 ). Generative theories redirected the attention of linguists and psycholinguists to syntax (and to a lesser extent, semantics). Rules were where the action was, because they seemed to best account for the productive and generative nature of linguistic knowledge. Words, on the other hand, were idiosyncratic (insofar as the mapping between meaning and phonological form was mostly arbitrary and variable across languages). They had to be learned, to be sure, but their seemingly unsystematic character suggested that learning had to be rote. The mental lexicon was a rather uninteresting place, necessary but rather dull.

In recent decades, words have made a comeback. Many linguists have come to see words not simply as flesh that gives life to grammatical structures, but as bones that are themselves grammatical rich entities. This sea change has accompanied the rise of usage-based theories of language (e.g., Langacker, 1987 ; Tomasello, 2003 ), which emphasize the context-sensitivity of word use. In some theories, the distinction between rule and word is blurred, with both seen as objects that implement form-mapping relationships ( Goldberg, 2003 ; Jackendoff, 2007 ). Within developmental psychology, words have always been of interest (after all, In the beginning, there was the word…) but more recent theories suggest that words may themselves be the foundational elements from which early grammar arises epiphenomenally ( Bates & Goodman, 1997 ; Tomasello, 2000 ). In the fields of psycholinguistics and computational linguistics, an explosion of findings indicate that interpretation of a sentence’s grammatical structure interacts with the comprehender’s detailed knowledge of properties of the specific words involved, the statistical patterns of usage, and that these interactions may occur at early stages of processing ( Altmann, 1998 ; Hare, McRae, & Elman, 2003 ; MacDonald, 1997 ; Roland & Jurafsky, 2002 ; Tanenhaus, Spivey-Knowlton, Eberhard, & Sedivy, 1995 ).

All of this has led to a sea change, resulting in the view of the mental lexicon as a data structure of tremendous richness and detail. And this, I want to propose, should begin to raise some worries. How much detail ought to go in the lexicon? Is there a principled way to adjudicate between information that belongs in the lexicon and information that belongs elsewhere?

In the remainder of this paper, I want to suggest that the metaphor of the mental lexicon as a dictionary-like data structure is not up to the job. But before going any further, I need to make something very clear. There is no question that lexical knowledge, that is, knowledge of words’ properties and appropriate usage, is extremely rich. To argue against the existence of a mental lexicon, which is what I will do, is not to argue that people lack knowledge of words. As a convenient way to talk about lexical knowledge, using the term ‘mental lexicon’ is not problematic. The issue I am concerned with is rather ‘What is the cognitive mechanism that encodes and deploys word knowledge?’ Is this knowledge really encoded in an enumerated declarative data structure, akin to a dictionary? If not, how then might lexical knowledge be instantiated? I will argue that the dictionary metaphor is ultimately infeasible, and will suggest an alternative.

What information might go in the mental lexicon?

The metaphor of the mental lexicon as a dictionary is pervasive and compelling. For many theories it not just a metaphor, but is taken seriously as a description of the data structure that stores knowledge about words. I take Jackendoff’s (2002) description of the lexicon as a reasonable and typical account:

For a first approximation, the lexicon is the store of words in long-term memory from which the grammar constructs phrases and sentences (p. 130)…[A lexical entry] lists a small chunk of phonology, a small chunk of syntax, and a small chunk of semantics (p. 131).

But just how small is the “small chunk” of phonology, syntax, and semantics? This is a crucial question. As already observed, the richness of lexical entries has grown considerably over past decades. Is there a principled limit? And are there empirical tests we might impose that would help adjudicate which information is likely to reside in a lexical entry, and which information might arise from other knowledge sources, including semantics, pragmatics, and general world knowledge? Let me turn now to knowledge of a specific lexical class of items, verbs, and discuss what recent experimental research might contribute to this discussion. I will begin by reviewing some of the findings in the sentence processing literature — particularly results dealing with verb behavior — because those have been especially significant in promoting the view of an enriched lexicon.

Sentence processing and the lexicon — background

Within the psycholinguistic literature, much of the data that led to new views about the lexicon have resulted not from the direct study of lexical representations per se, but as a by-product of a theoretical debate in recent decades regarding the mechanisms of sentence processing. The controversy has to do with how language users deal with the challenge of interpreting sentences that are presented in real time, incrementally, word by word. In many cases, this leads to points within the sentence that are at least temporarily ambiguous in the sense that they are compatible with very different grammatical structures and very different meaning interpretations. Usually (but not always), the ambiguities are eventually resolved by the remainder of the sentence. The question is how comprehenders deal with the temporary ambiguities at the point where they arise. Two major possibilities have been proposed.

One hypothesis has been that processing occurs in at least two stages (e.g., Frazier, 1978 , 1990 , 1995 ; Frazier & Rayner, 1982 ; Rayner, Carlson, & Frazier, 1983 ). During the first stage, the comprehender attempts to create a syntactic parse tree that best matches the input up to that point. It is assumed that in this first stage, only basic syntactic information regarding the current word is available, such as the word’s grammatical category and a limited set of grammatically relevant features. In the case of verbs, this information might include the verb’s selectional restrictions, subcategorization information, thematic roles, etc. ( Chomsky, 1965 , 1981 ; Dowty, 1991 ; Katz & Fodor, 1963 ). Then, at a slightly later point in time, a second stage of processing occurs in which more complete information about the lexical item becomes available, including the word’s semantic and pragmatic information, as well as world knowledge. Interpretive processes also operate, and these may draw on contextual information. Occasionally, the information that becomes available during this second pass might force a revision of the initial parse. However, if the heuristics are efficient and well motivated, this two-stage approach permits a quick and dirty analysis that will work most of the time without the need for revision.

The contrasting theory, often described as a constraint-based, probabilistic, or expectation-driven approach, emphasizes the probabilistic and context-sensitive aspects of sentence processing ( Altmann, 1998 , 1999 ; Altmann & Kamide, 1999 ; Elman, Hare, & McRae, 2005 ; Ford, Bresnan, & Kaplan, 1982 ; MacDonald, 1993 ; MacDonald, Pearlmutter, & Seidenberg, 1994 ; MacWhinney & Bates, 1989 ; McRae, Spivey-Knowlton, & Tanenhaus, 1998 ; St.John & McClelland, 1990 ; Tanenhaus & Carlson, 1989 ; Trueswell, Tanenhaus, & Garnsey, 1994 ). This approach assumes that comprehenders use all idiosyncratic lexical, semantic, and pragmatic information about each incoming word to determine a provisional analysis. Of course, temporary ambiguities in the input may still arise, and later information in the sentence might reveal that the initial analysis was wrong. Thus, both approaches need to deal with the problem of ambiguity resolution. The question is whether they make different predictions about processing that can be tested experimentally.

This debate has led to a fruitful line of research that focuses on cases in which a sentence is temporarily ambiguous and allows for two (or more) different structural interpretations. What is of interest is what happens when the ambiguity is resolved and it becomes clear which of the earlier possible interpretations is correct. The assumption is that if the sentence is disambiguated to reveal a different structure than the comprehender had assumed, there will be some impact on processing, either through an increased load resulting from recovery and reinterpretation, or perhaps simply as a result of a failed expectation.

Various measures have been used as markers of the processing effect that occurs at the disambiguation point in time, including reading times, patterns of eye movements, or EEG activity. These measures in turn provide evidence for how the comprehender interpreted the earlier fragment and therefore (a) what information was available at that time and (b) what processing strategy was used. Clearly, the many links in this chain form a valid argument only when all the links are well motivated; if any aspect of the argument is faulty, then the entire conclusion is undermined. It is not surprising that this issue has been so difficult to resolve to everyone’s satisfaction.

Over the years, however, the evidence in favor of the constraint-based, probabilistic approach has grown, leading many (myself included) to view this as the better model of human sentence processing. It is this research that has supported the enriched lexicon hypothesis. In what follows, I begin by describing several studies in which the results imply that a great deal of detailed and verb-specific information is available to comprehenders. Although first set of data are amenable to the strategy of an enriched lexicon, we quickly come upon data for which this is a much less reasonable alternative. These are the data that pose a dilemma for the lexicon.

Arguments for an enriched lexicon

The relationship between meaning and complement structure preferences.

One much studied structural ambiguity is that which arises at the postverbal noun phrase (NP) in sentences such as The boy heard the story was interesting. In this context, the story (at the point where it occurs) could either be the direct object (DO) of heard , or it could be the subject noun of a sentential complement (SC; as it ends up being in this sentence). The two-stage model predicts that the DO interpretation will be favored initially, even though hear admits both possibilities, and there is support for this prediction ( Frazier & Rayner, 1982 ). However, proponents of the constraint-based approach have pointed out that at least three other factors might be responsible for such a result: (1) the relative frequency that a given verb occurs with either a DO or SC ( Garnsey, Pearlmutter, Meyers, & Lotocky, 1997 ; Holmes, 1987 ; Mitchell & Holmes, 1985 ); (2) the relative frequency that a given verb takes an SC with or without the disambiguating but optional complementizer that ( Trueswell, Tanenhaus, & Kello, 1993 ); and (3) the plausibility of the postverbal NP as a DO for that particular verb ( Garnsey et al., 1997 ; Pickering & Traxler, 1998 ; Schmauder & Egan, 1998 ).

The first of these factors — the statistical likelihood that a verb appears with either a DO or SC structure — has been particularly perplexing. The prediction is that if comprehenders are sensitive to the usage statistics of different verbs, then when confronted with a DO/SC ambiguity, comprehenders will prefer the interpretation that is consistent with that verb’s bias. Some studies report either late or no effects of verb bias (e.g., Ferreira & Henderson, 1990 ; Mitchell, 1987 ). More recent studies, on the other hand, have shown that verb bias does affect comprehenders’ interpretation of such temporarily ambiguous sequences ( Garnsey et al., 1997 ; Trueswell et al., 1993 ; but see Kennison, 1999 ). Whether or not such information is used at early stages of processing is important not only because of its processing implications but because, if it is, this then implies that the detailed statistical patterns of subcategorization usage will need to be part of a verb’s lexical representation.

One possible explanation for the discrepant experimental data is that many of the verbs that show such DO/SC alternations have multiple senses, and these senses may have different subcategorization preferences ( Roland & Jurafsky, 1998 , 2002 ). This raises the possibility that a comprehender might disambiguate the same temporarily ambiguous sentence fragment in different ways, depending on the inferred meaning of the verb. That meaning might in turn be implied by the context that precedes the sentence. A context that primes the sense of the verb that more frequently occurs with DOs should generate a different expectation than a context that primes a sense that has an SC bias.

Hare, McRae, and Elman (2004 ; Hare et al., 2003 ) tested this possibility. Several large text corpora were analyzed to establish the statistical patterns of usage that were associated with verbs (DO vs. SC) and in which different preferences were found for different verb senses. The corpus analyses were used to construct pairs of two sentence stories; in each pair, the second target sentence contained the same verb in a sequence that was temporarily (up to the postverbal NP) ambiguous between a DO or SC reading. The first sentence provided a meaning biasing context. In one case, the context suggested a meaning for the verb in the target sentence that was highly correlated with a DO structure. In the other case, the context primed another meaning of the verb that occurred more frequently with an SC structure. Both target sentences were in fact identical till nearly the end. Thus, sometimes the ambiguity was resolved in a way that did not match participants’ predicted expectations. The data (reviewed in more detail in Hare, Elman, Tabaczynski, & McRae, in press ) suggest that comprehenders’ expectancies regarding the subcategorization frame in which a verb occurs is indeed sensitive to statistical patterns of usage that are associated not with the verb in general, but with the sense-specific usage of the verb. A computational model of these effects is described in Elman et al. (2005) .

A similar demonstration of the use of meaning to predict structure is reported in Hare et al. (in press) . That study examined expectancies that arise during incremental processing of sentences that involve verbs such as collect , which can occur in either a transitive construction (e.g., The children collected dead leaves , in which the verb has a causative meaning) or an intransitive construction (e.g., The rainwater collected in the damp playground , in which the verb is inchoative). Here again, at the point where the syntactic frame is ambiguous (at the verb, The children collected… or The dead leaves collected… ), comprehenders appeared to expect the construction that was appropriate given the likely meaning of the verb (causative vs. inchoative). In this case, the meaning was biased by having subjects that were either good causal agents (e.g., children in the first example above) or good themes ( rainwater in the second example).

These experiments suggest that the lexical representation of verbs must not simply include information regarding the verb’s overall structural usage patterns, but that this information regarding the syntactic structures associated with a verb is sense-specific, and a comprehender’s structural expectations are modulated by the meaning of the verb that is inferred from the context. This results in a slight enrichment of the verb’s lexical representation, but can be easily accommodated within the traditional lexicon.

Verb specific thematic role filler preferences

Another well studied ambiguity is that which arises with verbs such as arrest . These are verbs that can occur in both the active voice (as in The man arrested the burglar ) and in the passive (as in, The man was arrested by the policeman ). The potential for ambiguity arises because relative clauses in English ( The man who was arrested… ) may occur in a reduced form in which who was is omitted. This gives rise to The man arrested… , which is ambiguous. Until the remainder of the sentence is provided, it is temporarily unclear whether the verb is in the active voice (and the sentence might continue as in the first example) or whether this is the start of a reduced relative construction, in which the verb is in the passive (as in The man arrested by the policeman was innocent .)

In an earlier study, Taraban and McClelland (1988) found that when participants read sentences involving ambiguous prepositional attachments, e.g., The janitor cleaned the storage area with the broom… or The janitor cleaned the storage area with the solvent… , reading times were faster in sentences involving more typical fillers of the instrument role (in these examples, broom rather than solvent ). McRae et al. (1998) noted that in many cases, similar preferences appear to exist for verbs that can appear in either the active or passive voice. For many verbs, there are nominals that are better fillers of the agent role than the passive role, and vice versa.

This led McRae et al. (1998) to hypothesize that when confronted with a sentence fragment that is ambiguous between a Main Verb and Reduced Relative reading, comprehenders might be influenced by the initial subject NP and whether it is a more likely agent or patient. In the first case, this should encourage a Main Verb interpretation; in the latter case, a Reduced Relative should be favored. This is precisely what McRae et al. found to be the case. The cop arrested… promoted a Main Verb reading over a Reduced Relative interpretation, whereas The criminal arrested… , increased the likelihood of the Reduced Relative reading. McRae et al. concluded that the thematic role specifications for verbs must go beyond simple categorical information, such as Agent , Patient , Instrument , Beneficiary , etc. The experimental data suggest that the roles contain very detailed information about the preferred fillers of these roles, and that the preferences are verb-specific.

There is one additional finding that provides an important qualification of this conclusion. It turns out that different adjectival modifiers of the same noun can also affect its inferred thematic role. Thus, a shrewd, heartless gambler is a better agent of manipulate than a young, naïve gambler ; conversely, the latter is a better filler of the same verb’s patient role ( McRae, Ferretti, & Amyote, 1997 ). If conceptually based thematic role preferences are verb-specific, the preferences seem to be finer grained than simply specifying the favored lexical items that fill the role. Rather, the preferences may be expressed at the level of the semantic features and properties that characterize the nominal.

This account of thematic roles resembles that of Dowty (1991) in that both accounts suggest that thematic roles have internal structure. But the McRae et al. (1997 ; McRae et al., 1998 ) results further suggest a level of information that goes considerably beyond the limited set of proto-role features envisioned by Dowty. McRae et al. interpreted these role-filler preferences as reflecting comprehenders’ specific knowledge of the event structure associated with different verbs. This appeal to event structure, as we shall see below, will figure significantly in phenomena that are not as easily accommodated by the lexicon.

We have seen that verb-specific preferences for their thematic role fillers arise in the course of sentence processing. Might such preferences also be revealed in word-word priming? The answer is yes. Ferretti, McRae, and Hatherell (2001) found that verbs primed nouns that were good fillers for their agent, patient, or instrument roles. In a subsequent study, McRae, Hare, Elman, and Ferretti (2005) tested the possibility that such priming might go in the opposite direction, that is, that when a comprehender encounters a noun, the noun serves as a cue for the event in which it most typically participates, thereby priming verbs that describe that event activity. This prediction is consistent with literature on the multiple forms of organization of autobiographical event memory ( Anderson & Conway, 1997 ; Brown & Schopflocher, 1998 ; Lancaster & Barsalou, 1997 ; Reiser, Black, & Abelson, 1985 ). As predicted, priming was found.

The above experiments further extend the nature of the information that must be encoded in a verb’s lexical representation. In addition to sense-specific structural usage patterns, the verb’s lexical entry must also encode verb-specific information regarding the characteristics of the nominals that best fit that verb’s thematic roles.

The studies reviewed are but a few of very many similar experiments that have suggested that the lexical representation for verbs must include subentries about all the verb’s senses. Furthermore, for each sense, all possible subcategorization frames would be shown. For each verb-sense-subcategorization combination, additional information would be indicating the probability of each combination. Finally, similar information would be needed for every verb-sense-thematic role possibility. The experimental evidence indicates that in many cases, this latter information will be detailed, highly idiosyncratic of the verb, and represented at the featural level (e.g., Ferretti et al., 2001 ; McRae et al., 1997 ).

Challenges for the mental lexicon

Thus far, the experimental data suggest that comprehenders’ knowledge of fairly specific (and sometimes idiosyncratic) aspects of a verb’s usage is available and utilized early in sentence processing. This information includes sense-specific subcategorization usage patterns, as well as the properties of the nominals that are expected to fill the verb’s thematic roles. All of this expands the contents of the verb’s lexical representation, but not infeasibly so. We now turn to additional phenomena that will be problematic for the traditional view of the mental lexicon as an enumerative data structure, akin to a dictionary.

The effect of aspect

As noted above, Ferretti et al. (2001) found that verbs were able to prime their preferred agents, patients, and instruments. However, no priming was found from verbs to the locations in which their associated actions take place. Why might this be? One possibility is that locations are not as tightly associated with an event as are other participating elements. However, Ferretti, Kutas, and McRae (2007) noted that in that experiment the verb primes for locations were in the past tense ( e.g. , skated — arena ), and possibly interpreted by participants as having perfective aspect. Because the perfective signals that the event has concluded, it is often used to provide background information prefatory to the time period under focus (as in Dorothy had skated for many years and was now looking forward to her retirement ). Imperfective aspect, on the other hand, is used to describe events that are either habitual or on-going; this is particularly true of the progressive. Ferretti et al. hypothesized that although a past perfect verb did not prime its associated location, the same verb in the progressive might do so because of the location’s greater salience to the unfolding event.

This prediction was borne out. The two word prime had skated failed to yield significant priming for arena in a short SOA naming task, relative to an unrelated prime; but the two word prime was skating did significantly facilitate naming. In an ERP version of the experiment, the typicality of the location was found to affect expectations. Sentences such as The diver was snorkeling in the ocean (typical location) elicited lower amplitude N400 responses at ocean , compared to The diver was snorkeling in the pond at pond . The N400 is interpreted as an index of semantic expectancy, and the fact that typicality of agent-verb-location combinations affected processing at the location indicates that this information must be available early in processing.

The ability of verbal aspect to manipulate sentence processing by changing the focus on an event description can also be seen in the very different domain of pronoun interpretation. The question arises, How do comprehenders interpret a personal pronoun in one sentence when there are two potential referents in a previous sentence, and both are of the same gender (e.g., Sue disliked Lisa intensely. She _____). In this case, the reference is ambiguous.

One possibility is that there is a fixed preference, such that the pronoun is usually construed as referring to the referent that is in (for example) Subject position of the previous sentence. Another possibility, suggested by Kehler, Kertz, Rohde, and Elman (2008) is that pronoun interpretation depends on the inferred coherence relations between the two sentences ( Kehler, 2002 ). Under different discourse conditions, different interpretations might be preferred.

In a prior experiment, Stevenson, Crawley, and Kleinman (1994) asked participants to complete sentence pairs such as John handed a book to Bob. He ___ in which the pronoun could equally refer to either John (who in this context is said to fill the Source thematic role) or Bob (who fills the Goal role). Stevenson et al. found that Goal continuations (in which he is understood as referring to Bob ) and Source continuations ( he refers to John ) were about evenly split, 49%–51%. Kehler et al. (2008) suggested that, as was found in the Ferretti et al. (2007) study, aspect might alter this result. The reasoning was that perfective aspect tends to focus on the end state of an event, whereas imperfective aspect makes the on-going event more salient. When the event is construed as completed, the coherence of the discourse is most naturally maintained by continuing the story, what Kehler (2002) and Hobbs (1990) have called an Occasion coherence relation. Because continuations naturally focus on the Goal, the preference for Goal interpretations should increase. This appears to be the case. When participants were given sentences in which the verb was in the imperfective, such as John was handing a book to Bob , and then asked to complete a following sentence that began He ___, participants generated significantly more Source interpretations (70%) than for sentences in which the verb had perfective aspect. This result is consistent with the Ferretti et al. (2007) interpretation of their data, namely, that aspect alters the way comprehenders construe the event structure underlying an utterance. This in turn makes certain event participants more or less salient.

Let us return now to the effect of aspect on verb argument expectations. These results have two important implications. First, the modulating effect of aspect is not easily accommodated by spreading activation accounts of priming. In spreading activation models, priming is accomplished via links that connect related words and which serve to pass activation from one to another. These links are not thought to be subject to dynamic reconfiguration or context-sensitive modulation. In Section 4, I describe an alternative mechanism that might account for these effects.

The second implication has to do with how verb argument preferences are encoded. Critically, the effect seems to occur on the same time scale as other information that affects verb argument expectations (this was demonstrated by Experiment 3 in Ferretti et al. (2007 ), in which ERP data indicated aspectual differences within 400 ms of the expected word’s presentation). The immediate accessibility and impact of this information would make it a likely candidate for inclusion in the verb’s lexical representation. But logically, it is difficult to see how one would encode such a dynamic contingency on thematic role requirements.

Thus, although the patterns of ambiguity resolution described in earlier sections, along with parallel findings using priming ( Ferretti et al., 2001 ; McRae et al., 2005 ) might be accommodated by enriching the information in the lexical representations of verbs, the very similar effects of aspect do not seem amenable to a similar account. In languages such as English, a verb’s aspect is not an intrinsic property of the verb, yet the particular choice of aspect used in a given context affects expectations regarding the expectations regarding the verb’s arguments.

If verb aspect can alter the expected arguments for a verb, what else might do so? The concept of event representation has emerged as a useful way to understand several of the earlier studies. If we consider the question from the perspective of event representation, viewing the verb as providing merely some of the cues (albeit very potent ones) that tap into event knowledge, then several other candidates suggest themselves.

Dynamic alterations in verb argument expectations

If we think in terms of verbs as cues and events as the knowledge they target, then it should be clear that although the verb is obviously a very powerful cue, and that its aspect may alter the way the event is construed, there are other cues that change the nature of the event or activity associated with the verb. For example, the choice of agent of the verb may signal different activities. A sentence-initial noun phrase such as The surgeon… is enough to generate expectancies that constrain the range of likely events. In isolation, this cue is typically fairly weak and unreliable, but different agents may combine with the same verb to describe quite different events.

Consider the verb cut . Our expectations regarding what will be cut, given a sentence that begins The surgeon cuts… are quite different than for the fragment The lumberjack cuts… These differences in expectation clearly reflect our knowledge of the world. This is not remarkable. The critical question is, What is the status of such knowledge? No one doubts that a comprehender’s knowledge of how and what a surgeon cuts, versus what a lumberjack cuts, plays an important role in comprehension at some point.

The crucial question, for purposes of deciding what information is included in a lexical entry and what information arises from other knowledge sources, is when this knowledge enters into the unfolding process of comprehension. This is because timing has been an important adjudicator for models of processing and representation. If the knowledge is available very early — perhaps even immediately on encountering the relevant cues — then this is a challenge for two-stage serial theories (in which only limited lexical information is available during the first stage). Importantly, it is also problematic for standard theories of the lexicon.

Agent dependencies

Bicknell, Elman, Hare, McRae, and Kutas (2010) hypothesized that if different agent-verb combinations imply different types of events, this might lead comprehenders to expect different patients for the different events. This prediction follows from a study by Kamide, Altmann, and Haywood (2003) . Kamide et al. employed a paradigm in which participants’ eye movements toward various pictures were monitored as they heard sentences such as The man will ride the motorbike or The girl will ride the carousel (all combinations of agent and patient were crossed) while viewing a visual scene containing a man, a girl, a motorbike, a carousel, and candy. At the point when participants heard The man will ride… , Kamide et al. found that there were more looks toward the motorbike than to the carousel, and the converse was true for The girl will ride…. The Bicknell et al. study was designed to look specifically at agent-verb interactions and to see whether such effects also occurred during self-paced reading; and if so, how early in processing.

A set of verbs such as cut , save , and check were first identified as potentially describing different events depending on the agent of the activity, and in which the event described by the agent-verb combination would entail different patients. These verbs were then placed in sentences in which the agent-verb combination was followed either by the congruent patient, as in The journalist checked the spelling of his latest report… or in which the agent-verb was followed by an incongruent patient, as in The mechanic checked the spelling of his latest report… (all agents of the same verb appeared with all patients, and a continuation sentence followed that increased the plausibility of the incongruent events). Participants read the sentences a word at a time, using a self-paced moving window paradigm.

As predicted, there was an increase in reading times for sentences in which an agent-verb combination was followed by an incongruent (though plausible) patient. The slowdown occurred one word following the patient, leaving open the possibility that the expectation reflected delayed use of world knowledge. Bicknell et al. therefore carried out a second experiment using the same materials, but recording ERPs as participants read the sentences. The rationale for this was that ERPs provide a more precise and sensitive index of processing than reading times. Of particular interest was the N400 component, since this provides a good measure of the degree to which a given word is expected and/or integrated into the prior context. As predicted, an elevated N400 was found for incongruent patients.

The fact that what patient is expected may vary as a function of specific particular agent-verb combinations is not in itself surprising. What is significant is that the effect occurs at the earliest possible moment, at the patient that immediately follows the verb. The timing of such effects has in the past often been taken as indicative of an effect’s source. A common assumption has been that immediate effects reflect lexical or ‘first-pass’ processing, and later effects reflect the use of semantic or pragmatic information. In this study, the agent-verb combinations draw upon comprehenders’ world knowledge. The immediacy of the effect would seem to require either that this information must be embedded in the lexicon, or else that world knowledge must be able to interact with lexical knowledge more quickly than has often typically been assumed.

Instrument dependencies

Can other elements in a sentence affect the event type that is implied by the verb? Consider again the verb cut . The Oxford English Dictionary shows the transitive form of this verb as having a single sense. WordNet gives 41 senses. The difference is that WordNet ’s senses more closely correspond to what one might call event types, whereas the OED adheres to a more traditional notion of sense that is defined by an abstract core meaning that does not depend on context. Yet cutting activities in different contexts may involve quite different sets of agents, patients, instruments, and even locations. The instrument is likely to be a particularly potent constraint on the event type.

Matsuki et al. (in press) tested the possibility that the instrument used with a verb would cue different event schemas, leading to different expectations regarding the most likely patient. Using a self-paced reading format, participants read sentences such as Susan used the scissors to cut the expensive paper that she needed for her project , or Susan used the saw to cut the expensive wood… Performance on these sentences was contrasted with that on the less expected Susan used the scissors to cut the expensive wood… or Susan used the saw to cut the expensive paper…. As in the Bicknell et al. study, materials were normed to ensure that there were no direct lexical associations between instrument and patient. An additional priming study was carried out in which instruments and patients served as prime-target pairs; no significant priming was found between typical instruments and patients (e.g., scissors-paper ) versus atypical instruments and patients (e.g., saw-paper; but priming did occur for a set of additional items that were included as a comparison set). As predicted, readers showed increased reading times for the atypical patient relative to the typical patient. In this study, the effect occurred right at the patient, demonstrating that the filler of the instrument role for a specific verb alters the restrictions on the filler of the patient role.

Discourse dependencies

The problems for traditional lexical representations should start to be apparent. But there is one final twist. So far, we have seen that expectations regarding one of a verb’s arguments may be affected by how another of its arguments is realized. Is this effect limited to argument-argument interactions, or can discourse level context modulate argument expectations?

Race, Klein, Hare, and Tanenhaus (2008) took a subset of the sentences used in the Bicknell et al. (2010) experiment, in which different agent-verb combinations led to different predictions of the most likely patient. Race et al. then created stories that preceded the sentences, and in which the overall context strongly suggested a specific event that would involve actions that might or might not be typical for a given agent-verb combination. For example, although normally The shopper saved… and The lifeguard saved… lead to expectations of some amount of money and some of person , respectively, if the prior context indicates that there is a disaster occurring, or if there is a sale in progress, then this information might override the typical expectancies. That is exactly what Race et al. found. This leads to a final observation: A verb’s preferred patients do not depend solely on the verb, nor on the specific filler of the agent role, nor on the filler of the instrument role, but also on information from the broader discourse context. The specifics of the situation in which the action occurs matter.

Now let us see what all of this implies as far as the lexicon is concerned.

Lexical knowledge without a lexicon

Where does lexical knowledge reside.

The findings above strongly support the position that lexical knowledge is quite detailed, often idiosyncratic and verb specific, and to brought bear at the earliest possible stage in incremental sentence processing. The examples above focused on verbs, and the need to encode restrictions (or preferences) over the various arguments with which they may occur. Taken alone, those results might be accommodated by simply providing greater detail in lexical entries in the mental lexicon, as standardly conceived.

Where things get sticky is when one also considers what seems to be the ability of dynamic factors to significantly modulate such expectations. These include the verb’s grammatical aspect, the agent and instrument that are involved in the activity, and the overall discourse context. That these factors should play a role in sentence processing, at some point in time, is not itself surprising. The common assumption has been that such dynamic factors lie outside the lexicon. This is, for example, essentially the position outlined in J. D. Fodor (1995) : “We may assume that there is a syntactic processing module, which feeds into, but is not fed by, the semantic and pragmatic processing routines…syntactic analysis is serial, with back-up and revision if the processor’s first hypothesis about the structure turns out later to have been wrong” (p. 435). More pithily, the data do not accord with the “syntax proposes, semantics disposes” hypothesis ( Crain & Steedman, 1985 ). Thus, what is significant about the findings above is that the influence of aspect, agent, instrument, and discourse all occur within the same time frame that has been used operationally to identify information that resides in the lexicon. This is important if we are to have some empirical basis for deciding what goes in the lexicon and what does not.

All of this places us in the uncomfortable position of having to make some tough decisions.

One option would be to abandon any hope of finding any empirical basis for determining the contents of the mental lexicon. One might simply stipulate that some classes of information reside in the lexicon and others do not. This is not a desirable solution. Note that even within the domain of theoretical linguistics, there has been considerable controversy regarding what sort of information belongs in the lexicon, with different theories taking different and often mutually incompatible positions (contrast, among many other examples, Chomsky, 1965 ; J. A. Fodor, 2002 ; Haiman, 1980 ; Jackendoff, 1983 , 2002 ; Katz & Fodor, 1963 ; Lakoff, 1971 ; Langacker, 1987 ; Levin & Hovav, 2005 ; Weinreich, 1962 ). If we insist that the form of the mental lexicon has no consequences for processing, and exclude data of this type, this puts us in the awkward position where we have no behavioral way to evaluate different proposals.

A second option would be to significantly enlarge the format of lexical entries so that they accommodate all the above information. This would be a logical conclusion to the trend that has appeared not only in the processing literature (e.g., in addition to the studies cited above, Altmann & Kamide, 2007 ; Kamide, Altmann, et al., 2003 ; Kamide, Scheepers, & Altmann, 2003 ; van Berkum, Brown, Zwitserlood, Kooijman, & Hagoort, 2005 ; van Berkum, Zwitserlood, Hagoort, & Brown, 2003 ) but also many recent linguistic theories (e.g., Bresnan, 2006 ; Fauconnier & Turner, 2002 ; Goldberg, 2003 ; Lakoff, 1987 ; Langacker, 1987 ; though many or perhaps all of these authors might not agree with such a conclusion). The lexicon has become increasingly rich and detailed in recent years. Why impose arbitrary limits on its contents?

One problem is that the combinatoric explosion this entails would be significant. In fact, given the unbounded nature of discourse contexts, it is unclear that this is even feasible. But it also presents us with a logical conundrum: If all this information resides in the lexicon, is there then any meaningful distinction between the lexicon and other linguistic modules?

The third option is the most radical. It is to consider the possibility that lexical knowledge might be instantiated in a very different way than through a mental dictionary, and to find a computational mechanism that permits the sorts of complex interactions that appear to be required to use words appropriately.

An alternative to the mental lexicon as dictionary

The common factor in the studies described above was the ability of sentential elements to interact in real time to produce an incremental interpretation that guided expectancies about upcoming elements. These can be thought of as very powerful context effects that modulate the meaning that words have. Alternatively (but equivalently) one can view words not as elements in a data structure that must be retrieved from memory, but rather as stimuli that alter mental states (which arise from processing prior words) in lawful ways. In this view, words are not mental objects that reside in a mental lexicon. They are operators on mental states.

This scheme of things can be captured by a model that instantiates a dynamical system. The system receives inputs (words, in this case) over time. The words perturb the internal state of the system (we can call it the “mental state”) as they are processed, with each new word altering the mental state in some way.

Over the years, a number of connectionist models have been developed that illustrate ways in which context can influence processing in complicated but significant ways (e.g., among many others, McClelland & Rumelhart, 1981 ; McRae et al., 1998 ; Rumelhart, Smolensky, McClelland, & Hinton, 1988 ; Taraban & Mc- Clelland, 1988 ). There is also a rich literature in the use of dynamical systems to model cognitive phenomena (e.g., Smith & Thelen, 1993 ; Spencer & Schöner, 2003 ; Tabor & Tanenhaus, 2001 ; Thelen & Smith, 1994 ).

One very simple architecture that illustrates how one might model the context- dependent nature of lexical knowledge is a connectionist model known as a simple recurrent network (SRN; Elman, 1990 ), shown schematically in Figure 1 . Each rectangular box stands for some set of processing units (akin to abstract, highly simplified neurons in a real neural network) that connects to the units in other layers (shown as other rectangles). The arrows illustrate the flow of activation. What I have called the “mental state” of the system is the activation pattern that is present in the Hidden Layer at any given point in time. This internal state varies as a function of both its own prior internal state (this is the result of the feedback connections that allow the state at time t to feed into the state at time t+1 ) and the external input. In this model, the inputs correspond to words, which might be represented either abstractly as binary valued vectors but could also be represented as phonological forms. Finally, at each point in time, the network produces some output on its Output Layer.

Simple Recurrent Network. Each layer is composed of one or more units. Information flows from input to hidden to output layers. In addition, at every time step t, the hidden unit layer receives input from the context layer, which stores the hidden unit activations from time t – 1.

Knowledge in such a network is contained in the pattern of connections between processing units, and in particular, in the strength of the weights. Prior to learning, the SRN’s weights are initialized with small random values. At this point in time, it knows nothing. Learning is done by example. Rather than instructing the network on explicit rules, the network is shown examples of well formed stimuli (in this case, grammatical sentences). In this example, the network was presented with a large number of sentences that exemplify the ways in which verb arguments may depend on complex interactions between each other.

One task that is deceptively simple but turns out to be very powerful is prediction. In this task, the network is presented with the words in a sentence in succession. At every time step, the network is asked to predict what the next word will be. Words are represented as arbitrary binary vectors, which deprives the network of prior information regarding the lexicosemantic properties of words. A simple learning algorithm ( Rumelhart, Hinton, & Williams, 1986 ) is used to gradually adjust the connection weights so that, over time, the network’s actual output more closely appropriates the desired output (in this case, the correct next word).

If the training data are sufficiently large and complex, the network will typically not be able to memorize the sentences. Given the nondeterministic nature of most sentences, this means that the network will not be able to literally predict successive words. What we really would hope for, and what the network succeeds in doing, is to learn the context-contingent dependencies that make some words more probable successors than others, and rules out some words as ungrammatical.

For example, after learning, and given the test sentence The girl ate the… , the network will not predict a single word, but all possible words that are sensible in this context, given the language sample it has experienced. Thus, it might predict sandwich , taco , cookie , and other edible nominals. Words that are either ungrammatical (e.g., verbs) or semantically or pragmatically inappropriate (e.g., rock ) will not be predicted.

From this behavior, we might infer the network learns the lexicogrammatical categories implicit in the training data. We can verify this by analyzing the internal representations that the network has learned for each word. These are instantiated in the activation patterns at the Hidden Layer level that are produced as the network incrementally processes successive words in sentences. These activations patterns are vectors, have a geometric interpretation as points in a high dimensional “mental space” of the network. The patterns have a similarity structure which corresponds to each word’s proximity to every other word in that space.

Figure 2 displays a hierarchical clustering tree that depicts that similarity structure. Words whose internal representations are close in the “mental space” are shown as leaves that are close on the tree. One can see through the structure of the tree that the network has grouped nouns apart from verbs, and that also makes finer grained distinctions within these two categories (e.g., animate vs. inanimate nouns; large animals vs. small animals; transitive vs. intransitive verbs , etc.). Figure 3 shows, in simplified cartoon form and in just three dimensions, what the actual spatial relations between the words’ internal representations might look like.

Hierarchical clustering diagram of hidden unit activation patterns in response to different words. The similarity between words and groups of words is reflected in the tree structure; items that are closer are joined lower in the tree.

Schematic visualization, in 3-D, of the high dimensional state space described by the SRN’s hidden unit layer. The state space is partitioned into different regions that correspond to grammatical and semantic catgories. Nesting relationship in space (e.g., HUMAN within ANIMATE within NOUN categories) reflect hierarchical relationships between categories.

Not visible in either of these representations is the trajectory or path over time through the network’s internal space that results when successive words in a sentence are presented. These trajectories reflect the intrinsic dynamics of the network, such that only some paths through state space are felicitous and well formed. Indeed, the network dynamics encode what we would conventionally think of as the grammar that underlies the language sample. An important discovery in recent years is that networks of this sort can implement recursive relationships that allow them to represent abstract long distance dependencies, and that the grammars that are learned generalize beyond the training data ( Boden & Blair, 2003 ; Boden & Wiles, 2000 ; Rodriguez, 2001 ; Rodriguez & Elman, 1999 ; Rodriguez, Wiles, & Elman, 1999 ). We will shortly see that such trajectories also play a role in encoding lexical knowledge.

Let us now consider a simple model that is exposed to language data in which knowledge of the proper use of a verb involves learning the complex dependencies between the specific arguments and adjuncts that may be used with the verb in different situations. Figure 4 is a schematic depiction of a family of sentences that illustrate such possibilities for the verb cut .

The verb cut may denote many different types of activities, depending on (among other factors) the agent, instrument, or location with which the verb occurs. These dependencies then affect what is the likely filler of the theme role for the verb.

After training, the network is then tested by inputting, a word at a time, various sentences that illustrate these complexities. The network’s predictions closely accord with what is appropriate, such that after processing The butcher uses a saw to cut… the network predicts that the next word will be meat , whereas after A person uses a saw to cut… the response is a tree. At a behavioral level, then, the network demonstrates that it has learned the lexical and grammatical regularities underlying the sentences.

How does the network do this? There are two interdependent strategies. First, the network learns to partition its internal representational space so that the internal representations that arise in real time as the network processes words reflect the basic lexico-semantic properties of the vocabulary (in much the same way as shown in Figure 3 ). Second, the syntagmatic knowledge governing argumentadjunct- verb interactions arise from the dynamical properties of the network (encoded in the weights between units). This means that when a word is processed, that word’s impact on the internal state combines with the prior context to generate predictions about the class of grammatically appropriate continuations would be. We can see this by plotting the trajectories over time for various sentences involving the same verb but different arguments or adjuncts.

Figure 5 . shows the trajectories through the network’s internal state space as it processes two sentences involving the verb cut . We see that the states for cut in the two instances are close in space, reflecting the fact that there is considerable overlap in the context specific meaning associated with these two usages. But the states are not identical, illustrating the ways in which the contingencies between a verb and the other elements in a sentence combine to determine (in this case) the likely filler of the theme role. This interaction between context and lexical knowledge is inextricable and immediate.

Trajectories through 3 (of 20 total) dimensions of an SRN’s hidden layer. These correspond to movement through the state space as the network processes the sentences “A person uses a saw to cut a tree” and “A butcher uses a saw to cut meat.” The state of the network resulting from any given word is what encodes its expectancies of what will follow. Thus, the states at “cut” in the two sentences differ, reflecting different expectations regarding the likely patient this is to follow (resulting from the use of different instruments). Once the patient is processed, it produces a state appropriate to the end of the sentence. This is why both patients produce very similar states.

Although the possibility of lexical knowledge without a lexicon might seem odd, the core ideas that motivate this proposal are not new. Many elements appear elsewhere in the literature. These include the following.

• The meaning of a word is rooted in our knowledge of both the material and the social world. The material world includes the world around us as we experience it (it is embodied), possibly indirectly. The social world includes cultural habits and artifacts; in many cases, these habits and artifacts have significance only by agreement (they are conventionalized). Similar points have been made by many others, notably including Wittgenstein (1966) , Hutchins (1994) and Fauconnier (1997 ; Fauconnier & Turner, 2002 ).
• Context is always with us. The meaning of a word is never “out of context”, although we might not always know what the context is (particularly if we fail to provide one). This point has been made by many, including Kintsch (1988) , Langacker (1987) , McClelland, St. John, & Taraban (1989) , and van Berkum et al. (2003 ; 2005) . This insight is of course also what underlies computational models of meaning that emphasize multiple co-occurrence constraints between words in order to represent them as points in a high dimensional space, such as LSA ( Landauer & Dumais, 1997 ), HAL ( Burgess & Lund, 1997 ), or probabilistic models ( Griffiths & Steyvers, 2004 ). The dynamical approach here also emphasizes the time course of processing that results from the incremental nature of language input.
• The drive to predict is a simple behavior with enormously important consequences. It is a powerful engine for learning, and provides important clues to latent abstract structure (as in language). Prediction lays the groundwork for learning about causation. These points have been made elsewhere by many, including Elman (1990) , Kahneman and Tversky (1973) , Kveraga, Ghuman, and Bar (2007) , Schultz, Dayan, and Montague (1997) , and Spirtes, Glymour, and Scheines (2000) . It should not be surprising that prediction would also be exploited for language learning and play a role in on-line language comprehension.
• Events play a major role in organizing our experience. Event knowledge is used to drive inference, to access memory, and affects the categories we construct. An event may be defined as a set of participants, activities, and outcomes that are bound together by causal interrelatedness. An extensive literature argues for this, aside from the studies described here, including work by Minsky (1974) , Schank and Abelson (1977) , and Zacks and Tversky (2001) ; see also Shipley and Zacks (2008) for a comprehensive collection on the role of event knowledge in perception, action, and cognition.
• Dynamical systems provide a powerful framework for understanding biologically based behavior. The nonlinear and continuous valued nature of dynamical systems allows them to respond in a graded manner under some circumstances, while in other cases their responses may seem more binary. Dynamical analyses figure prominently in the recent literature in cognitive science, including work by Smith and Thelen (1993 , 2003 ; Thelen & Smith, 1994 ), Spencer and Schöner (2003) , Spivey (2007) , Spivey and Dale (2004) , Tabor (2004) , Tabor, Juliano, and Tanenhaus (1997) and Tabor and Tanenhaus (2001) .

It must be emphasized that the model in Figure 6 is far too simple to serve as anything but a conceptual metaphor. It is intended to help visualize how the knowledge that we are removing from word-as-operand is moved into the processing mechanism on which word-as-operator acts. Many important details are omitted. Critically, this simple model is disembodied; it lacks the conceptual knowledge about events that comes from direct experience. The work described here has emphasized verbal language, and this model only captures the dynamics of the linguistic input. In a full model, one would want many inputs, corresponding to the multiple modalities in which we experience the world. Discourse involves many other types of interactions. For example, the work of Clark, Goldin- Meadow, McNeil, and many others makes it clear that language is well and rapidly integrated with gesture ( Clark, 1996 , 2003 ; Goldin-Meadow, 2003 ; McNeil, 1992 , 2005 ). The dynamics of such a system would be considerably more complex than those shown in Figure 6, since each input domain has its own properties and domain internal dynamics. In a more complete model, these would exist as coupled dynamical subsystems that interact.

How does this view affect the way we do business (or at least, study words)? Although I have argued that much of the behavioral phenomena described above are not easily incorporated into a mental lexicon as traditionally conceived, I cannot at this point claim that accommodating them in some variant of the lexicon is impossible. A parallel architecture of the sort described by Jackendoff (2002) , for example, if it permitted direct and immediate interactions among the syntactic, semantic, and pragmatic components of the grammar, might be able to account for the data described earlier. The important question would still remain about how to motivate what information is placed where, but these concerns do not in themselves rule out a lexical solution. Unfortunately, it is also then not obvious whether tests can be devised to distinguish between these proposals. This remains an open question for the moment.

However, theories can also be evaluated for their ability to offer new ways of thinking about old problems, or to provoke new questions that would not be otherwise asked. A theory might be preferred over another because it leads to a research program that is more productive than the alternative. Let me suggest two positive consequences to the sort of words-as-cues dynamical model I am outlining.

The first has to do with the role that theories play in the phenomena they predict. The assumption that only certain information goes in the lexicon, and that the lexicon and other knowledge sources respect modular boundaries with limited and late occurring interactions, drives a research program that discourages looking for evidence of richer and more immediate interactions. For example, the notion that selectional restrictions might be dynamic and context-sensitive is fundamentally not an option within the Katz and Fodor framework (1963) . The words-as-cues approach, in contrast, suggests that such interdependencies should be expected. Indeed, there should be many such interactions among lexical knowledge, context, and nonlinguistic factors, and these might occur early in processing. Many researchers in the field have already come to this point of view. It is a conclusion that, despite considerable empirical evidence, has been longer in the coming than it might have, given a different theoretical perspective.

A second consequence of this perspective is that it encourages a more unified view of phenomena that are often treated ( de facto , if not in principle) as unrelated. Syntactic ambiguity resolution, lexical ambiguity resolution, pronoun interpretation, text inference, and semantic memory (to chose but a small subset of domains) are studied by communities that do not always communicate well, and researchers in these areas are not always aware of findings from other areas. Yet these domains have considerable potential for informing each other. That is because, although they ultimately draw on a common conceptual knowledge base, that knowledge base can be accessed in different ways, and this in turn affects what is accessed. Consider how our knowledge of events might be tapped in a priming paradigm, compared with a sentence processing paradigm. Because prime-target pairs are typically presented with no discourse context, one might expect that a transitive verb prime might evoke a situation in which the fillers of both its agent and patient roles are equally salient. Thus, arresting should prime cop (typical arrestor) and also crook (typical arrestee). Indeed, this is what happens ( Ferretti et al., 2001 ). Yet this same study also demonstrated that when verb primes were embedded in sentence fragments, the priming of good agents or patients was contingent on the syntactic frame within which the verb occurred. Primes of the form She arrested the… facilitated naming of crook , but not cop . Conversely, the prime She was arrested by the… facilitated naming of cop rather than crook.

These two results demonstrate that although words in isolation can serve as cues to event knowledge, they are only one such cue. The grammatical construction within which they occur provides independent evidence regarding the roles played by different event participants ( Goldberg, 2003 ). And of course, the discourse context may provide further constraints on how an event is construed. Thus, as Race et al. (2008) found, although shoppers might typically save money and lifeguards save children, in the context of a disaster, both agents will be expected to save children.

There is a second consequence to viewing linguistic and nonlinguistic cues as tightly coupled. This has to do with learning and the problem of learnability. Much has been made about the so-called poverty of the stimulus ( Chomsky, 1980 , p. 34; Crain, 1991 ). The claim is that the linguistic data that are available to the child are insufficient to account for certain things that the child eventually knows about language. Two interesting things can be said about this claim. First, the argument typically is advanced “in principle” with scant empirical evidence that it truly is a problem. A search of the literature reveals a surprisingly small number of specific phenomena for which the poverty of the stimulus is alleged. Second, whether or not the stimuli available for learning are impoverished depend crucially on what one considers to be the relevant and available stimuli, and what the relevant and available aspects or properties of those stimuli are.

Our beliefs about what children hear seem to be based partly on intuition, partly on very small corpora, and partly on limited attempts to see whether children are in fact prone to make errors in the face of limited data. In at least some cases, more careful examination of the data and of what children do and can learn given those data do not support the poverty of the stimulus claim ( Ambridge, Pine, Rowland, & Young, 2008 ; Pullum & Scholz, 2002 ; Reali & Christiansen, 2005 ; Scholz & Pullum, 2002 ). It is not always necessary to see X in the input to know that X is true. It may be that Y and Z logically make X necessary ( Lewis & Elman, 2001 ).

If anything is impoverished, it is not the stimuli but our appreciation for how rich the fabric of experience is. The usual assumption is that the relevant stimuli consist of the words a child hears, and some of the arguments that have been used in support of the poverty of the stimulus hypothesis (e.g., Gold, 1967 ) have to do with what are essentially problems in learning syntactic patterns from positive only data. We have no idea how easy or difficult language learning is if the data include not only the linguistic input but also the simultaneous stream of nonlinguistic information that accompanies it. However, there are many examples that demonstrate that learning in one modality can be facilitated by use of information from another modality (e.g., Ballard & Brown, 1993 ; de Sa, 2004 ; de Sa & Ballard, 1998 ). Why should this not also be true for language learning as well?

Eliminating the lexicon is indeed radical surgery, and it is an operation that at this point many will not agree to. At the very least, however, I hope that by demonstrating that lexical knowledge without a lexicon is possible, others will be encouraged to seek out additional evidence for ways in which the many things that language users know is brought to bear on the way language is processed.

Acknowledgments

This work was supported by NIH grants HD053136 and MH60517 to Jeff Elman, Mary Hare, and Ken McRae; NSERC grant OGP0155704 to KM; NIH Training Grant T32-DC000041 to the Center for Research in Language (UCSD); and by funding from the Kavli Institute of Brain and Mind (UCSD). Much of the experimental work reported here and the insights regarding the importance of event knowledge in sentence processing are the fruit of a long-time productive collaboration with Mary Hare and Ken McRae. I am grateful to them for many stimulating discussions and, above all, for their friendship. I am grateful to Jay McClelland for the many conversations we have had over three decades. His ideas and perspectives on language, cognition, and computation have influenced my thinking in many ways. Finally, a tremendous debt is owed to Dave Rumelhart, whose suggestion (1979) that words do not have meaning, but rather that they are cues to meaning, inspired the proposal outlined here.

• Altmann GTM. Ambiguity in sentence processing. Trends in Cognitive Sciences. 1998; 2 :146–152. [ PubMed ] [ Google Scholar ]
• Altmann GTM. Thematic role assignment in context. Journal of Memory & Language. 1999; 41 (1):124–145. [ Google Scholar ]
• Altmann GTM, Kamide Y. Incremental interpretation at verbs: restricting the domain of subsequent reference. Cognition. 1999; 73 (3):247–264. [ PubMed ] [ Google Scholar ]
• Altmann GTM, Kamide Y. The real-time mediation of visual attention by language and world knowledge: Linking anticipatory (and other) eye movements to linguistic processing. Journal of Memory & Language. 2007; 57 (4):502–518. [ Google Scholar ]
• Ambridge B, Pine JM, Rowland CF, Young CR. The effect of verb semantic class and verb frequency (entrenchment) on children’s and adults’ graded judgements of argument-structure. Cognition. 2008; 106 (1):87–129. [ PubMed ] [ Google Scholar ]
• Anderson SJ, Conway MA. Representations of autobiographical memories. In: Conway MA, editor. Cognitive Models of Memory. Cambridge, MA: MIT Press; 1997. pp. 217–246. [ Google Scholar ]
• Ballard DH, Brown CM. Principles of animate vision. In: Aloimonos Y, editor. Active perception. Hillsdale, NJ: Lawrence Erlbaum Associates; 1993. pp. 245–282. [ Google Scholar ]
• Bates E, Goodman JC. On the inseparability of grammar and the lexicon: Evidence from acquisition, aphasia, and real-time processing. Language & Cognitive Processes. 1997; 12 :507–584. [ Google Scholar ]
• Bicknell K, Elman JL, Hare M, McRae K, Kutas M. Effects of event knowledge in processing verbal arguments. Journal of Memory & Language. 2010; 63 (4):489–505. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Boden M, Blair A. Learning the dynamics of embedded clauses. Applied Intelligence. 2003; 19 (1–2):51–63. [ Google Scholar ]
• Boden M, Wiles J. Context-free and context-sensitive dynamics in recurrent neural networks. Connection Science: Journal of Neural Computing, Artificial Intelligence & Cognitive Research. 2000; 12 (3–4):197–210. [ Google Scholar ]
• Bresnan J. Is syntactic knowledge probabilistic? Experiments with the English dative alternation. In: Featherston S, Sternefeld W, editors. Roots: Linguistics in Search of its Evidential Base. Berlin: Mouton de Gruyter; 2006. [ Google Scholar ]
• Brown NR, Schopflocher D. Event clusters: An organization of personal events in autobiographical memory. Psychological Science. 1998; 9 :470–475. [ Google Scholar ]
• Burgess C, Lund K. Modeling parsing constraints with high-dimensional context space. Language & Cognitive Processes. 1997; 12 :177–210. [ Google Scholar ]
• Chomsky N. Syntactic structures. The Hague: Mouton; 1957. [ Google Scholar ]
• Chomsky N. Aspects of the theory of syntax. Oxford, England: M.I.T. Press; 1965. [ Google Scholar ]
• Chomsky N. Rules and Representations. New York: Columbia University Press; 1980. [ Google Scholar ]
• Chomsky N. Lectures on Government and Binding. New York: Foris; 1981. [ Google Scholar ]
• Clark HH. Using Language. Cambridge: Cambridge University Press; 1996. [ Google Scholar ]
• Clark HH. Pointing and placing. In: Kita S, editor. Pointing. Where Language, Culture, and Cognition Meet. Hillsdale, NJ: Lawrence Erlbaum Associates; 2003. pp. 243–268. [ Google Scholar ]
• Crain S. Language acquisition in the absence of experience. Brain & Behavioral Sciences. 1991; 14 :597–611. [ Google Scholar ]
• Crain S, Steedman M. On not being led up the garden path: The use of context by the psychological parser. In: Dowty D, Karttunen L, Zwicky A, editors. Natural Language Processing: Psychological, Computational, and Theoretical Perspectives. Cambridge: Cambridge University Press; 1985. pp. 320–358. [ Google Scholar ]
• de Sa VR. Sensory modality segregation. In: Thurn S, Saul L, Schoelkopf B, editors. Advances in Neural Information Processing Systems. Vol. 16. Cambridge, MA: MIT Press; 2004. pp. 913–920. [ Google Scholar ]
• de Sa VR, Ballard DH. Category learning through multimodality sensing. Neural Computation. 1998; 10 (5):1097–1117. [ PubMed ] [ Google Scholar ]
• Dowty D. Thematic proto-roles and argument selection. Language. 1991; 67 :547–619. [ Google Scholar ]
• Elman JL. Finding Structure in Time. Cognitive Science. 1990; 14 (2):179–211. [ Google Scholar ]
• Elman JL, Hare M, McRae K. Cues, constraints, and competition in sentence processing. In: Tomasello M, Slobin D, editors. Beyond nature-nurture: Essays in honor of Elizabeth Bates. Mahwah, NJ: Lawrence Erlbaum Associates; 2005. pp. 111–138. [ Google Scholar ]
• Fauconnier G. Mappings in thought and language. New York, NY: Cambridge University Press; 1997. [ Google Scholar ]
• Fauconnier G, Turner M. The way we think: conceptual blending and the mind’s hidden complexities. New York: Basic Books; 2002. [ Google Scholar ]
• Ferreira F, Henderson JM. Use of verb information in syntactic parsing: Evidence from eye movements and word-by-word self-paced reading. Journal of Experimental Psychology: Learning, Memory, & Cognition. 1990; 16 (4):555–568. [ PubMed ] [ Google Scholar ]
• Ferretti TR, Kutas M, McRae K. Verb Aspect and the Activation of Event Knowledge. Journal of Experimental Psychology: Learning, Memory, & Cognition. 2007; 33 (1):182–196. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Ferretti TR, McRae K, Hatherell A. Integrating verbs, situation schemas, and thematic role concepts. Journal of Memory & Language. 2001; 44 (4):516–547. [ Google Scholar ]
• Fodor JA. The lexicon and the laundromat. In: Merlo P, Stevenson S, editors. The Lexical Basis of Sentence Processing. Amsterdam: John Benjamins; 2002. pp. 75–94. [ Google Scholar ]
• Fodor JD. Thematic roles and modularity. In: Altmann GTM, editor. Cognitive Models of Speech Processing. Cambridge, MA: MIT Press; 1995. pp. 434–456. [ Google Scholar ]
• Ford M, Bresnan J, Kaplan RM. A competence-based theory of syntactic closure. In: Bresnan J, editor. The Mental Representation of Grammatical Relations. Cambridge: MIT Press; 1982. pp. 727–796. [ Google Scholar ]
• Frazier L. Unpublished PhD. University of Connecticut; 1978. On comprehending sentences: Syntactic parsing strategies. [ Google Scholar ]
• Frazier L. Parsing modifiers: Special-purpose routines in the human sentence processing mechanism. Hillsdale, NJ: Erlbaum; 1990. [ Google Scholar ]
• Frazier L. Constraint satisfaction as a theory of sentence processing. Journal of Psycholinguistic Research Special Issue: Sentence processing: I. 1995; 24 (6):437–468. [ PubMed ] [ Google Scholar ]
• Frazier L, Rayner K. Making and correcting errors during sentence comprehension: Eye movements in the analysis of structurally ambiguous sentences. Cognitive Psychology. 1982; 14 (2):178–210. [ Google Scholar ]
• Garnsey SM, Pearlmutter NJ, Meyers E, Lotocky MA. The contribution of verbbias and plausibility to the comprehension of temporarily ambiguous sentences. Journal of Memory & Language. 1997; 37 :58–93. [ Google Scholar ]
• Gold EM. Language identification in the limit. Information & Control. 1967; 16 :447–474. [ Google Scholar ]
• Goldberg AE. Constructions: a new theoretical approach to language. Trends in Cognitive Sciences. 2003; 7 (5):219–224. [ PubMed ] [ Google Scholar ]
• Goldin-Meadow S. Hearing gesture: How our hands help us think. Cambridge, MA: MIT Press; 2003. [ Google Scholar ]
• Griffiths TL, Steyvers M. A probabilistic approach to semantic representation. Paper presented at the Proceedings of the 24th Annual Conference of the Cognitive Science Society; George Mason University; 2004. [ Google Scholar ]
• Haiman J. Dictionaries and encyclopedias. Lingua. 1980; 50 :329–357. [ Google Scholar ]
• Hare M, Elman JL, Tabaczynski T, McRae K. The wind chilled the spectators but the wine just chilled: Sense, structure, and sentence comprehension. Cognitive Science in press. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Hare M, McRae K, Elman JL. Sense and structure: Meaning as a determinant of verb subcategorization preferences. Journal of Memory & Language. 2003; 48 (2):281–303. [ Google Scholar ]
• Hare M, McRae K, Elman JL. Admitting that admitting verb sense into corpus analyses makes sense. Language & Cognitive Processes. 2004; 19 :181–224. [ Google Scholar ]
• Hobbs JR. Literature and cognition. Stanford, CA: Center for the Study of Language and Information; 1990. [ Google Scholar ]
• Holmes VM. Syntactic parsing: In search of the garden path. In: Coltheart M, editor. Attention and performance 12: The psychology of reading. Hove, England UK: Lawrence Erlbaum Associates; 1987. pp. 587–599. [ Google Scholar ]
• Hutchins E. Cognition in the World. Cambridge, MA: MIT Press; 1994. [ Google Scholar ]
• Jackendoff R. Semantics and cognition. Cambridge, MA: MIT Press; 1983. [ Google Scholar ]
• Jackendoff R. Foundations of Language: Brain, Meaning, Grammar, and Evolution. Oxford: Oxford University Press; 2002. [ PubMed ] [ Google Scholar ]
• Jackendoff R. A parallel architecture perspective on language processing. Brain Research. 2007; 1146 :2–22. [ PubMed ] [ Google Scholar ]
• Kahneman D, Tversky A. On the psychology of prediction. Psychological Review. 1973; 80 (4):237–251. [ Google Scholar ]
• Kamide Y, Altmann GTM, Haywood SL. The time-course of prediction in incremental sentence processing: Evidence from anticipatory eye movements. Journal of Memory & Language. 2003; 49 (1):133–156. [ Google Scholar ]
• Kamide Y, Scheepers C, Altmann GTM. Integration of syntactic and semantic information in predictive processing: Cross-linguistic evidence from German and English. Journal of Psycholinguistic Research. 2003; 32 (1):37–55. [ PubMed ] [ Google Scholar ]
• Katz JJ, Fodor JA. The structure of a semantic theory. Language. 1963; 39 (2):170–210. [ Google Scholar ]
• Kehler A. Coherence, reference, and the theory of grammar. Palo Alto, CA: CSLI Publications, University; 2002. [ Google Scholar ]
• Kehler A, Kertz L, Rohde H, Elman JL. Coherence and coreference revisited. Journal of Semantics. 2008; 25 :1–44. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Kennison SM. American English usage frequencies for noun phrase and tensed sentence complement-taking verbs. Journal of Psycholinguistic Research. 1999; 28 (2):165–177. [ PubMed ] [ Google Scholar ]
• Kintsch W. The role of knowledge in discourse comprehension: A construction-integration model. Psychological Review. 1988; 95 (2):163–182. [ PubMed ] [ Google Scholar ]
• Kveraga K, Ghuman AS, Bar M. Top-down predictions in the cognitive brain. Brain & Cognition. 2007; 65 (2):145–168. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Lakoff G. Presuppositions and relative well-formedness. In: Steinberg D, Jakobovitz L, editors. Semantics. London: Cambridge University Press; 1971. pp. 329–340. [ Google Scholar ]
• Lakoff G. Women, fire, and dangerous things: What categories reveal about the mind. Chicago: University of Chicago Press; 1987. [ Google Scholar ]
• Lancaster JS, Barsalou L. Multiple organisations of events in memory. Memory. 1997; 5 :569–599. [ PubMed ] [ Google Scholar ]
• Landauer TK, Dumais ST. A solution to Plato’s problem: The Latent Semantic Analysis theory of the acquisition, induction, and representation of knowledge. Psychological Review. 1997; 104 :211–240. [ Google Scholar ]
• Langacker RW. Foundations of Cognitive Grammar. Vol. 1. Stanford: Stanford University Press; 1987. [ Google Scholar ]
• Levin B, Hovav MR. Argument realization. Cambridge: Cambridge University Press; 2005. [ Google Scholar ]
• Lewis JD, Elman JL. A connectionist investigation of linguistic arguments from the poverty of the stimulus: Learning the unlearnable. In: Moore JD, Stenning K, editors. Proceedings of the Twenty-third Annual Conference of the Cognitive Science Society. Mahwah, NJ: Lawrence Erlbaum Associates; 2001. pp. 552–557. [ Google Scholar ]
• MacDonald MC. The interaction of lexical and syntactic ambiguity. Journal of Memory & Language. 1993; 32 (5):692–715. [ Google Scholar ]
• MacDonald MC. Lexical representations and sentence processing: An introduction. Language & Cognitive Processes. 1997; 12 (2&3):121–136. [ Google Scholar ]
• MacDonald MC, Pearlmutter NJ, Seidenberg MS. The lexical nature of syntactic ambiguity resolution. Psychological Review. 1994; 101 (4):676–703. [ PubMed ] [ Google Scholar ]
• MacWhinney B, Bates E. The Crosslinguistic Study of Sentence Processing. Cambridge; New York: Cambridge University Press; 1989. [ Google Scholar ]
• Matsuki K, Chow T, Hare M, Elman JL, Scheepers C, McRae K. Dynamic combination of concepts during on-line language comprehension. Journal of Experimental Psychology: Learning, Memory and Cognition in press. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• McClelland JL, Rumelhart DE. An interactive activation model of context effects in letter perception: Part 1. An account of basic findings. Psychological Review. 1981; 86 :287–330. [ PubMed ] [ Google Scholar ]
• McClelland JL, St John MF, Taraban R. Sentence comprehension: A parallel distributed processing approach. Language & Cognitive Processes. 1989; 4 :287–336. [ Google Scholar ]
• McNeil D. Hand and mind: What gestures reveal about thought. Chicago: University of Chicago Press; 1992. [ Google Scholar ]
• McNeil D. Gesture and thought. Chicago: University of Chicago Press; 2005. [ Google Scholar ]
• McRae K, Ferretti TR, Amyote L. Thematic roles as verb-specific concepts. Language and Cognitive Processes: Special Issue on Lexical Representations in Sentence Processing. 1997; 12 :137–176. [ Google Scholar ]
• McRae K, Hare M, Elman JL, Ferretti TR. A basis for generating expectancies for verbs from nouns. Memory & Cognition. 2005; 33 (7):1174–1184. [ PubMed ] [ Google Scholar ]
• McRae K, Spivey-Knowlton MJ, Tanenhaus MK. Modeling the influence of thematic fit (and other constraints) in on-line sentence comprehension. Journal of Memory & Language. 1998; 38 (3):283–312. [ Google Scholar ]
• Minsky M. A framework for representing knowledge. Massachusetts Institute of Technology; 1974. [ Google Scholar ]
• Mitchell DC. Lexical guidance in human parsing: Locus and processing characteristics. In: Coltheart M, editor. Attention and Performance XII: The psychology of reading. Hillsdale, NJ: Erlbaum; 1987. pp. 601–618. [ Google Scholar ]
• Mitchell DC, Holmes VM. The role of specific information about the verb in parsing sentences with local ambiguity. Journal of Memory & Language. 1985; 24 :542–559. [ Google Scholar ]
• Pickering MJ, Traxler MJ. Plausibility and recovery from garden-paths: An eyetracking study. Journal of Experimental Psychology: Learning, Memory, & Cognition. 1998; 24 :940–961. [ Google Scholar ]
• Pullum GK, Scholz BC. Empirical assessment of stimulus poverty arguments. Linguistic Review. 2002; 19 (1–2):9–50. [ Google Scholar ]
• Race D, Klein N, Hare M, Tanenhaus MK. What do shoppers expect to save? Agents influence patients via events, not associative priming. Poster presented at the 20th Annual CUNY Conference on Human Sentence Processing; Chapel Hill, NC. 2008. [ Google Scholar ]
• Rayner K, Carlson M, Frazier L. The interaction of syntax and semantics during sentence processing. Journal of Verbal Learning & Verbal Behavior. 1983; 22 :358–374. [ Google Scholar ]
• Reali F, Christiansen MH. Uncovering the Richness of the Stimulus: Structure Dependence and Indirect Statistical Evidence. Cognitive Science: A Multidisciplinary Journal. 2005; 29 (6):1007–1028. [ PubMed ] [ Google Scholar ]
• Reiser BJ, Black JB, Abelson RP. Knowledge structures in the organization and retrieval of autobiographical memory. Cognitive Psychology. 1985; 17 :89–137. [ Google Scholar ]
• Rodriguez P. Simple recurrent networks learn context-free and context-sensitive languages by counting. Neural Computation. 2001; 13 (9):2093–2118. [ PubMed ] [ Google Scholar ]
• Rodriguez P, Elman JL. Watching the transients: viewing a simple recurrent network as a limited counter. Behaviormetrika. 1999; 26 (1):51–74. [ Google Scholar ]
• Rodriguez P, Wiles J, Elman JL. A recurrent neural network that learns to count. Connection Science. 1999; 11 (1):5–40. [ Google Scholar ]
• Roland D, Jurafsky D. How verb subcategorization frequencies are affected by corpus choice. Paper presented at the Proceedings of COLING-ACL; 1998; Montreal, Canada. 1998. [ Google Scholar ]
• Roland D, Jurafsky D. Verb sense and verb subcategorization probabilities. In: Merlo P, Stevenson S, editors. The Lexical Basis of Sentence Processing: Formal, Computational, and Experimental Issues. Amsterdam: John Benjamins; 2002. pp. 325–347. [ Google Scholar ]
• Rumelhart DE. Some problems with the notion that words have literal meanings. In: Ortony A, editor. Metaphor and Thought. Cambridge: Cambridge University Press; 1979. pp. 71–82. [ Google Scholar ]
• Rumelhart DE, Hinton GE, Williams RJ. Learning internal representations by error propagation. In: Rumelhart DE, McClelland JL, editors. Parallel Distributed Processing: Explorations in the Microstructure of Cognition. Vol. 2. Cambridge, MA: MIT Press; 1986. pp. 318–362. [ Google Scholar ]
• Rumelhart DE, Smolensky P, McClelland JL, Hinton GE. Schemata and sequential thought processes in PDP models. In: McClelland JL, Rumelhart DE, editors. Parallel Distributed Processing: Explorations in the Microstructure of Cognition. Vol. 2. Cambridge, MA: MIT Press; 1988. pp. 7–57. [ Google Scholar ]
• Schank RC, Abelson RP. Scripts, plans, goals, and understanding: An inquiry into human knowledge structures. Hillsdale, NJ: Lawrence Erlbaum Associates; 1977. [ Google Scholar ]
• Schmauder AR, Egan MC. The influence of semantic fit on on-line sentence processing. Memory & Cognition. 1998; 26 (6):1304–1312. [ PubMed ] [ Google Scholar ]
• Scholz BC, Pullum GK. Searching for arguments to support linguistic nativism. Linguistic Review. 2002; 19 (1–2):185–223. [ Google Scholar ]
• Schultz W, Dayan P, Montague PR. A neural substrate of prediction and reward. Science. 1997; 275 (5306):1593–1599. [ PubMed ] [ Google Scholar ]
• Shipley TF, Zacks JM, editors. Understanding Events: How Humans See, Represent, and Act on Events. Oxford: Oxford University Press; 2008. [ Google Scholar ]
• Smith LB, Thelen E. A Dynamic Systems Approach to Development: Applications. Cambridge, MA: MIT Press; 1993. [ Google Scholar ]
• Smith LB, Thelen E. Development as a dynamic system. Trends in Cognitive Sciences. 2003; 7 (8):343–348. [ PubMed ] [ Google Scholar ]
• Spencer JP, Schöner G. Bridging the representational gap in the dynamic systems approach to development. Developmental Science. 2003; 6 (4):392–412. [ Google Scholar ]
• Spirtes P, Glymour CN, Scheines R. Causation, Prediction, and Search. Cambridge, MA: MIT Press; 2000. [ Google Scholar ]
• Spivey MJ. The Continuity of Mind. Oxford: Oxford University Press; 2007. [ Google Scholar ]
• Spivey MJ, Dale R, editors. On the continuity of mind: Toward a dynamical account of cognition. San Diego, CA: Elsevier Academic Press; 2004. [ Google Scholar ]
• St John M, McClelland JL. Learning and applying contextual constraints in sentence comprehension. Artificial Intelligence. 1990; 46 :217–257. [ Google Scholar ]
• Stevenson RJ, Crawley RA, Kleinman D. Thematic roles, focus and the representation of events. Language & Cognitive Processes. 1994; 9 (4):519–548. [ Google Scholar ]
• Tabor W. Effects of merely local syntactic coherence on sentence processing. Journal of Memory & Language. 2004; 50 :355–370. [ Google Scholar ]
• Tabor W, Juliano C, Tanenhaus MK. Parsing in a dynamical system: An attractorbased account of the interaction of lexical and structural constraints in sentence processing. Language & Cognitive Processes. 1997; 12 (2&3):211–271. [ Google Scholar ]
• Tabor W, Tanenhaus MK. Dynamical systems for sentence processing. In: Christiansen MH, Chater N, editors. Connectionist psycholinguistics. Westport, CT: Ablex Publishing; 2001. pp. 177–211. [ Google Scholar ]
• Tanenhaus MK, Carlson G. Lexical structure and language comprehension. Cambridge, MA: MIT Press; 1989. [ Google Scholar ]
• Tanenhaus MK, Spivey-Knowlton MJ, Eberhard KM, Sedivy JC. Integration of visual and linguistic information in spoken language comprehension. Science. 1995; 268 (5217):1632–1634. [ PubMed ] [ Google Scholar ]
• Taraban R, McClelland JL. Constituent attachment and thematic role assignment in sentence processing: Influences of content-based expectations. Journal of Memory & Language. 1988; 27 (6):597–632. [ Google Scholar ]
• Thelen E, Smith LB. A dynamic Systems Approach to the Development of Cognition and Action. Cambridge, MA: MIT Press; 1994. [ Google Scholar ]
• Tomasello M. The item based nature of children’s early syntactic development. Trends in Cognitive Science. 2000; 4 :156–163. [ PubMed ] [ Google Scholar ]
• Tomasello M. Constructing a language: A Usage-Based Theory of Language Acqusition. Cambridge, MA: Harvard University Press; 2003. [ Google Scholar ]
• Trueswell JC, Tanenhaus MK, Garnsey SM. Semantic influences on parsing: Use of thematic role information in syntactic ambiguity resolution. Journal of Memory & Language. 1994; 33 (3):285–318. [ Google Scholar ]
• Trueswell JC, Tanenhaus MK, Kello C. Verb-specific constraints in sentence processing: Separating effects of lexical preference from garden-paths. Journal of Experimental Psychology: Learning, Memory, & Cognition. 1993; 19 (3):528–553. [ PubMed ] [ Google Scholar ]
• van Berkum JJA, Brown CM, Zwitserlood P, Kooijman V, Hagoort P. Anticipating upcoming words in discourse: Evidence from ERPs and reading times. Journal of Experimental Psychology: Learning, Memory and Cognition. 2005; 31 (3):443–467. [ PubMed ] [ Google Scholar ]
• van Berkum JJA, Zwitserlood P, Hagoort P, Brown CM. When and how do listeners relate a sentence to the wider discourse? Evidence from the N400 effect. Cognitive Brain Research. 2003; 17 (3):701–718. [ PubMed ] [ Google Scholar ]
• Weinreich U. Lexicographic definition in descriptive semantics. In: Householder FW, Saporta W, editors. Problems in lexicography. Vol. 21. Bloomington, IN: Indiana University Research Center in Anthropology, Folklore, and Linguistics; 1962. [ Google Scholar ]
• Wittgenstein L. Philosophical investigations. Oxford, England: Macmillian; 1966. [ Google Scholar ]
• Zacks JM, Tversky B. Event structure in perception and conception. Psychological Bulletin. 2001; 127 (1):3–21. [ PubMed ] [ Google Scholar ]