Elsevier

Cognition

Volume 87, Issue 2, March 2003, Pages 101-127
Cognition

On the processing of regular and irregular forms of verbs and nouns: evidence from neuropsychology

https://doi.org/10.1016/s0010-0277(02)00200-7Get rights and content

Abstract

Following acquired brain damage, a native English speaking patient (AW) encountered problems accessing phonology in speech production, while her ability to access word meaning appeared to be intact. In a series of tasks, AW was presented either with a verb, and was asked to produce its past tense or past participle (walk → “walked”), or with a noun, and was asked to produce its plural (glove → “gloves”). A stark dissociation was found: while AW responded accurately with regular forms of verbs (walked) and nouns (gloves), performance was significantly less accurate with irregular forms (found; children). The appearance of a selective deficit for irregular forms in conditions of impaired lexical access is in line with dual-mechanism accounts, which proposes that irregular forms are specified in the lexicon whereas regular forms are computed via rule-based mechanisms. In contrast, AW's data are problematic for connectionist accounts that do not posit separate mechanisms for processing regular and irregular forms, including the connectionist model recently proposed by Joanisse and Seidenberg (Proceedings of the National Academy of Sciences USA 96 (1999) 7592) which successfully simulated a variety of earlier neuropsychological findings. Analyses of AW's responses shed light on further details of the representation and processing of regular and irregular inflected forms.

Introduction

In many languages, speakers modify the meaning of a word by changing the suffix appended to the end of the word. In English, for example, information about number (plural/singular) is conveyed by the presence/absence of the suffix -s at the end of nouns, and the suffixes -s, -ing, and -ed at the end of verbs express when the action or state described by the verbs takes place. Suffixation is an extremely productive process that speakers extend to recently introduced words like fax (faxes) and e-mail (e-mailed). But there are exceptions. English provides illustrative cases of such exceptions: a few plural nouns are not produced by adding the suffix -s (teeth, women, fish) and a good number of verbs take a past tense form that does not contain the suffix -ed (ran, sat, went). The occurrence of these irregular forms raises interesting questions about the process of word formation: are the mechanisms for word formation the same for regular and irregular words? If the mechanisms are not the same, how do they differ? Not only do the answers to these questions elucidate the organization of word formation processing (morphology) but they also have important implications for our understanding of the structure and functioning of the mental dictionary (the lexicon).

For the past 20 years, the debate about regular and irregular form processing has focused on English past tense inflection, which has become a crucial test case for theories of word formation. A view that has found wide support in linguistics and psycholinguistics is the dual-mechanism account, which holds that different mechanisms are at play with regularly inflected verbs (walkwalked, argueargued) and irregularly inflected verbs (runran, sitsat) (Bauer, 1983, Marslen-Wilson and Tyler, 1997, Pinker, 1991, Ullman, 2001). For regular past tenses, rule-based mechanisms add the suffix -ed to the verb stem (walk+ed → walked).1 Irregular past tenses are not obtained via rule-based mechanisms, but are stored in the lexicon; consequently, to produce these irregular forms speakers have to access the lexicon. Because the process for producing regular and irregular past tenses follows different principles, it is likely that different areas in the brain support the processing of these verbs (on this point see e.g. Marslen-Wilson & Tyler, 1998).

A contrasting view holds that regular and irregular past tenses are computed by a single mechanism. This view was implemented in a number of connectionist simulations (e.g. MacWhinney and Leinbach, 1991, Marchman, 1993, Plunkett and Marchman, 1993, Rumelhart and McClelland, 1986). These simulations operate on the basis of associative mechanisms established through learning which link input nodes specifying the phonology of a verb stem (e.g. walk) to output units specifying the phonology of the verb's past tense (e.g. walked). The theoretical import of these simulations is twofold: they demonstrate that it is not necessary to postulate different mechanisms for regular and irregular past tenses, nor is it necessary to postulate rule-based mechanisms for the formation of regular past tenses.

In support of either of these views, researchers have cited various sorts of data: results from reaction time experiments, computer simulations, normal and abnormal language acquisition, and historical linguistics (for a review, see Clahsen, 1999, Marcus, 2000, Pinker, 1999, Ullman, 2001). Recently, in an attempt to gather data that may resolve the current debate as well as shed light on the brain mechanisms underlying word processing, researchers have turned their attention to brain-damaged patients with acquired language disorders. Ullman et al. (1997) tested patients with language impairments due to different pathologies in a past tense generation task. Patients were presented with a sentence like “Every day I dig a hole. Just like every day, yesterday I ____ a hole” and asked to complete the sentence by providing the past tense form of the verb, “dug” in this example. Discrepancies were observed in patients' abilities to produce regular vs. irregular past tenses. Aphasic patients with posterior lesions and word-finding problems along with a group of patients diagnosed with Alzheimer's disease performed worse with irregularly inflected verbs. A contrasting pattern was reported for one aphasic patient with an anterior lesion and for a group of patients with Parkinson's disease; they encountered greater problems producing regular as opposed to irregular past tenses. Moreover, these patients performed differently than normal controls when asked to generate the past tense of novel verbs (vask, tunch): while normal controls typically added the suffixed -ed (“vasked”, “tunched”), the patients did so far less frequently. Similar dissociations have now been observed in a variety of tasks. Marslen-Wilson and Tyler, 1997, Marslen-Wilson and Tyler, 1998 have documented selective deficits in the comprehension of regular or irregular past tenses with the priming paradigm. Four patients with acquired language deficits heard two words (a prime and a target) and decided whether the second word corresponded to a familiar word. Various prime–target pairs were used: morphologically related (jumpedjump), semantically related (swangoose), and unrelated (lockedjump). For two patients, priming (faster responses than those for unrelated pairs) appeared in the presence of pairs formed by semantically related words and by irregular past tenses (foundfind). For two other patients, facilitation only appeared in the presence of regular past tenses (walkedwalk). Similar dissociations have been observed in reading (Ullman et al., 1997, Ullman et al., in press), and in a judgment task in which patients were asked to rate the ‘goodness’ of correct and incorrect past tense forms (e.g. dug/digged, rob/rob; Ullman et al., in press). In sum, the dissociation documented between regular and irregular past tenses both in speech production and comprehension has been interpreted as supporting the principle claim of the dual-mechanism account: that regular and irregular forms are processed by distinct (and neuroanatomically segregated) mechanisms (see also Bullinaria & Chater, 1995). The neuropsychological data have been considered to be incompatible with the view, endorsed by single-mechanism theories and implemented in connectionist networks, that regularly and irregularly inflected verbs recruit identical processes.

Joanisse and Seidenberg (1999) challenged the claim that the neuropsychological evidence is irreconcilable with models that do not incorporate distinct processes for regular and irregular inflections. They implemented a connectionist model that simulated different tasks with past tenses, including the elicitation task devised by Ullman et al. (1997). The model includes units devoted to the encoding of verb meaning2 and verb phonology, respectively (see Fig. 1). The model also incorporates separate phonological units for speech input and speech output. A key feature of the model is that regular and irregular past tenses are processed similarly, in the sense that identical mechanisms are implicated in the processing of both classes of verbs. This feature does not mean that semantic and phonological information are equally critical for the production of regular, irregular, and novel past tense forms. Because a novel verb like wug does not have any meaning, semantic information cannot contribute to the generation of the past tense form of wug, which has to be derived “by analogy” from the phonology of known verbs. In contrast, semantic information is crucial for irregularly inflected verbs; the generation of their past tense form depends on the establishment of a link between their semantic representation and their representations in both input and output phonology. In this way, the past tense of irregularly inflected verbs will not conform to the statistically dominant pattern. To simulate the neuropsychological data, Joanisse and Seidenberg (1999) selectively damaged different components of their connectionist model. Damage to the semantic units impaired the generation of regular, irregular, and novel past tenses, but the largest impairment appeared with irregular forms. Damage to the phonological units also affected the processing of all three types of verbs,3 but novel verbs were impacted most severely. Joanisse and Seidenberg (1999) were therefore able to reproduce (at least part of) the dissociations reported in neuropsychology with regular, irregular, and novel past tenses. The fact that these dissociations emerged in a connectionist network that does not explicitly incorporate specific mechanisms for regular and irregular inflections calls into question the claim that the neuropsychological data are incompatible with this sort of model. Of course, the ability to replicate the neuropsychological dissociation comes with an added cost: semantics is called upon for the processing of the regular/irregular verb distinction that has traditionally been considered an idiosyncrasy of language with relevance only for language processing. By contrast, the dual-mechanism account supports the opposing view that the regular/irregular verb distinction is confined to the realm of language processing.

The computational simulations reported by Joanisse and Seidenberg's (1999) also have implications for our understanding of the deficits that have been observed in neuropsychology with English past tenses. In particular, deficits selectively or more severely impacting irregular past tense processing should be associated with damaged semantics. The expectation is then to find a deficit for irregular past tenses in patients with impaired semantics, but not in patients with impaired phonology. Indeed, Joanisse and Seidenberg (1999) speculated that the deficit for irregular inflections that Ullman et al. (1997) reported in their patients might have resulted from a semantic impairment. These patients present with posterior aphasia and Alzheimer's disease, cognitive deficits that are frequently accompanied by semantic impairment. The consequences of damaged semantics have been assessed by Patterson, Lambon Ralph, Hodges, and McClelland (2001) in a group of patients with semantic dementia, a form of progressive memory and language disorder that affects semantic knowledge in particular. Patients performed significantly worse with irregular verbs than regular verbs in the past tense generation task (Ullman et al., 1997) and in a recognition task in which patients chose the correct form between two alternatives (bought/buyed, saved/sove). Patients' accuracy in both tasks correlated with their knowledge of verb meaning as shown by a synonym judgment task in which patients were asked to identify verbs with similar meanings (e.g. for grind the correct response would be crush, not sip). Taken together, the Patterson et al. (2001) data suggest a causal link between a semantic deficit and a deficit for irregular inflections and thus provide support to the Joanisse and Seidenberg's (1999) account. This conclusion is strengthened by the observation that the two patients reported by Marslen-Wilson and Tyler, 1997, Marslen-Wilson and Tyler, 1998 with a selective deficit for irregular past tenses did not show semantic priming, a finding which leads one to suspect the co-occurrence of a semantic deficit in these patients. Finally, further converging evidence comes from a study conducted by Tyler, deMornay-Davies et al. (2002) on four patients with semantic deficits due to left inferior temporal gyrus damage following herpes simplex encephalitis. As a group, the patients were impaired in their ability to produce irregular past tenses in an elicitation task similar to that of Ullman et al. (1997).

A selective deficit for irregular past tenses receives a different explanation within the dual-mechanism account: this deficit is likely to follow a problem in processing lexical information (instead of a semantic impairment). It should be recalled that in the dual-mechanism account, the phonological forms of irregularly inflected words are stored in the lexicon, and problems in retrieving the phonological forms from the lexicon should lead to difficulties particularly with irregularly inflected words. In theory, there are multiple ways in which the dual-mechanism models could account for the co-occurrence of deficits for semantic knowledge and irregular past tenses. Marslen-Wilson and Tyler (1998) provided an articulate account of why, in a comprehension task, semantic processing and the processing of irregular past verbs were both impaired. At base, dual-mechanism models are committed to the prediction that patients with selective problems of lexical access should fail with irregular verbs.

Anomia, an acquired speech deficit characterized by patients' difficulties finding the correct word, offers a further opportunity to test the dual-mechanism account and the single-mechanism account implemented by Joanisse and Seidenberg (1999). The speech of anomics is fluent and (at least in some patients) grammatical, though punctuated by frequent pauses in which they struggle to find the desired word (Allport, 1983, Garrett, 1992, Geshwind, 1967, Kay and Ellis, 1987). Anomics' speech failures relate to a problem of retrieving the word phonology stored in the lexicon, either because this information has been lost or is not accessible. Anomics may occasionally succeed in reporting fragmentary information about the sound of the target word – its onset phoneme(s), or words that sound like the target word (see e.g. Henaff Gonon, Bruckert, & Michel, 1989) – but in many cases the word phonology remains unavailable (see e.g. Badecker, Miozzo, & Zanuttini, 1995). Information about the syntactic features of words that is stored in the lexicon (grammatical class, number, verb aspect, etc.) may or may not remain available to anomic patients. In some anomics, semantic processing appears to be intact, as demonstrated by the fact that, for example, the patients can provide accurate descriptions or synonyms of the words they fail to produce. The dissociation found in anomia between intact semantic processing and impaired phonological retrieval is of particular relevance here.

In production tasks in which verb stems have to be retrieved from the lexicon (e.g. in naming depicted actions) anomics should be equally impaired with regular and irregular verbs, as long as these verbs are matched for variables known to affect anomics' naming (e.g. word frequency). However, differences between regular and irregular forms are expected in the stem → past tense generation task (Ullman et al., 1997). More importantly, the dual-mechanism account and the Joanisse and Seidenberg's (1999) account make contrasting predictions about how patients with anomia (but no semantic damage) will perform in tasks of regular and irregular past tense production. The dual-mechanism account anticipates a deficit in the production of irregular past tenses. Because anomics have problems retrieving the forms stored in the lexicon, and because the forms of irregular past tenses have to be retrieved from the lexicon, the generation of irregular past tenses should be impaired. As long as rule-based mechanisms are intact, and the -ed suffix can be retrieved successfully, anomics should be able to add the -ed suffix to the stem, and thereby respond correctly in instances in which they are asked to produce regularly inflected verbs and novel verbs. With respect to the Joanisse and Seidenberg's (1999) model, their simulations of phonological damage recreate the conditions observed in forms of anomia not associated with a semantic impairment; the expectation is to replicate in anomia the pattern that is observed in the model with damaged phonology. Namely, we should find that anomics are, relative to normal controls, impaired with regular, irregular, and novel past tenses. Moreover, anomics should perform equally well with regular and irregular past tenses, and better with these verbs than with novel verbs.

Patients classified as anomics were reported by Ullman et al., 1997, Ullman et al., in press, but unfortunately this study did not include analyses of patients' semantic processing. Currently we do not know how anomics with spared semantics process regular, irregular, and novel past tenses. It is this issue that is addressed in the present paper. Here I report an in-depth investigation of AW, a brain-damaged English-speaking patient with acquired anomia, and describe her ability to process regular, irregular, and novel verb inflections. Because English also includes nouns that are irregular in their plural form (e.g. women, deer, and oxen), AW's production of regular and irregular plural nouns was also examined to see whether similar deficits emerge with nouns and verbs. The presentation of the patient's data is divided into two parts: the first part documents AW's word-finding difficulties and shows that AW's semantic processing is intact; the second part describes how AW produces inflected forms of familiar and novel verbs, and nouns.

Section snippets

Patient description

AW is a native English speaker and a housewife with a high-school education. She is right handed with no familial sinistrality. The patient reported not having close relatives suffering from degenerative dementia, schizophrenia or developmental learning disorders such as dyslexia. At the time of testing she was 71, and had suffered a stroke 2 years earlier. A brain CT scan taken 3 days after her stroke showed lesions of the basal ganglia, the frontal white matter, and of the medial and superior

Part 1: lexical vs. semantic impairment

The results presented in this section establish that the word-finding difficulties observed in AW arise as a consequence of her problems retrieving word lexical phonology, that is information about the phonology of the word that is stored within the lexicon. Such information specifies the phonemes and other features of the word (e.g. stress, number of syllables) that are critical for the correct realization of the word. By contrast, phonological processes that take place after the retrieval of

Part 2: generation of regular vs. irregular forms

This section presents several tasks that compare AW's ability to generate regular and irregular verbs and nouns. The issue is whether AW, a patient with problems in retrieving word lexical phonology, shows selective deficits in producing regular or irregular forms. Another task requires AW to generate the past tense of novel verbs and is intended to determine whether the patient applies regular inflections to these verbs.

General discussion

AW is a patient with acquired word-finding difficulties caused by a left-temporal lesion. Detailed analyses indicate that AW's naming impairment is due to a problem of retrieving word phonology. When the patient failed to produce a word, very rarely could she retrieve fragments of the sound of the recalcitrant word; by contrast, AW seemed able to access precise semantic descriptions of objects as evidenced by her excellent performance even in demanding semantic tasks with nouns and verbs. Of

Acknowledgements

This research was supported by a Keck Foundation Grant. I thank Diana Van Lanker for having referred AW to our laboratory, Janelle Barnes and Melissa Jacobs for their careful preparation and administration of the test materials, Lice Ghilardi for the examination of the patient's neuroradiological records, and Brenda Rapp, Uri Hadar, and three anonymous reviewers for their comments on an earlier version of the manuscript. Most of all, I am indebted to patient AW: her continued patience and good

References (46)

  • L Bauer

    English word-formation

    (1983)
  • J.A Bullinaria et al.

    Connectionist modeling: implications for cognitive neuropsychology

    Language and Cognitive Processes

    (1995)
  • H Clahsen

    Lexical entries and rules of language: a multidisciplinary study of German inflections

    Behavioral and Brain Sciences

    (1999)
  • N.W Francis et al.

    Frequency analysis of English usage

    (1982)
  • M.A Gernsbacher

    Resolving 20 years of inconsistent interaction between lexical familiarity and orthography, concreteness, and polysemy

    Journal of Experimental Psychology: General

    (1984)
  • N Geshwind

    The nature of naming errors

    Cortex

    (1967)
  • H Goodglass et al.

    The assessment of aphasia and related disorders

    (1972)
  • M Halle et al.

    Distributed morphology and the pieces of inflections

  • M Henaff Gonon et al.

    Lexicalization in an anomic patient

    Neuropsychologia

    (1989)
  • A.E Hillis et al.

    Cognitive and neural mechanisms underlying visual and semantic processing: implications from “optic aphaisa”

    Journal of Cognitive Neuroscience

    (1995)
  • J.R Hodges et al.

    Semantic dementia: progressive fluent aphasia with temporal lobe atrophy

    Brain

    (1992)
  • M.F Joanisse et al.

    Impairments in verb morphology after brain injury: a connectionist model

    Proceedings of the National Academy of Sciences USA

    (1999)
  • J Kay et al.

    A cognitive neuropsychological case study of anomia

    Brain

    (1987)
  • Cited by (81)

    • Morphological segmentation of nonwords in individuals with acquired dyslexia

      2022, Journal of Memory and Language
      Citation Excerpt :

      Some individuals with aphasia (typically those with agrammatic speech and left frontal lesions; see Ullman et al., 2005) have been found to have an impairment in reading aloud regular English inflected forms compared with irregular inflected forms (e.g., Badecker & Caramazza, 1987, 1991; Hamilton & Coslett, 2008; Marin, Saffran, & Schwartz, 1976; Ullman et al., 1997; Ullman, Hickok, & Pinker, 1995). In contrast, other aphasics (typically those with word-finding difficulties and left temporal/temporo-parietal lesions; see Ullman et al., 2005) have been shown to have a deficit in reading irregular inflections compared with regular inflections (e.g., Miozzo, 2003; Ullman et al., 2005). What these prior findings seem to suggest is that the processing of affixed words (including regular inflections) is handled by a mechanism that is clearly distinct from the process of retrieving monomorphemic words and irregularly inflected words from the lexicon.

    • Morphological derivation overflow as a result of disruption of the left frontal aslant white matter tract

      2015, Brain and Language
      Citation Excerpt :

      Thus, the patient could be simply tired or resource depleted during task execution during surgery and chose to respond employing the more automatic route (the default morphological derivation rule application). Nevertheless this account would not explain why both greater impairment of lexical retrieval forms and also greater rule-based impairments have been reported after brain lesions (de Diego-Balaguer, 2004; Miozzo, 2003; Miozzo et al., 2010), and in our patient, would not account for the spared irregular inflection despite the lexical retrieval deficit in derivation. To conclude, to the best of our knowledge, this is the first study reporting relation between verb generation and intraoperative stimulation of the left FAT.

    • Inflectional morphology in primary progressive aphasia: An elicited production study

      2014, Brain and Language
      Citation Excerpt :

      To our knowledge, the neural correlates of deficits in inflectional morphology in PPA have not been systematically investigated. Neuropsychological studies in other patient cohorts have provided some evidence suggesting that deficits in regular morphology are associated with frontal and basal ganglia damage, in contrast to deficits in irregular morphology, which are related to temporal lobe lesions (Marin, Saffran, & Schwartz, 1976; Miozzo, 2003; Tyler et al., 2002; Ullman et al., 2005). A number of neuroimaging studies in healthy controls have attempted to identify brain regions differentially involved in regular or irregular morphology, yet findings have been inconsistent (Jaeger et al., 1996; Ullman et al., 1997; see Desai, Conant, Waldron, & Binder, 2006 for review).

    • Lesions impairing regular versus irregular past tense production

      2013, NeuroImage: Clinical
      Citation Excerpt :

      All items were monosyllabic in their present and past-tense forms. Verb item set 1 (stem task) consisted of 28 regular and 28 irregular verb stimuli from Woollams et al. (Miozzo, 2003). Verb item set 2 (sentence frame task) consisted of 27 regular and 27 irregular verbs.

    View all citing articles on Scopus
    View full text