Aphasia is a term used to describe a family of language disorders in which lexical and sentence-level dysfunctions result from damage to specific areas of the brain. It is commonly linked to brain damage following a stroke, head injury, tumor, or disease, though stroke is the most common cause.[1] The first three months after the damaging event is the usual window for "spontaneous recovery;" in general, recovery from traumatic brain injury progresses more quickly than that from stroke, while hemmorhagic stroke recovery occurs more easily than recovery from thrombo-embolic stroke.[2]

Types of Aphasia with Regard to the Brain

This section provides only a most cursory overview of the types of aphasia as related to specific areas of the human brain. See references for sources that address the psychological, psycholinguistic, and neuroscience basis for aphasia studies.

There are competing theories regarding the language centers of the brain and how trauma may affect such areas. In general, the current understanding of aphasia and brain injury suggests that lesions or damaged regions of the the language areas as well as the non-language sections of the brain can produce aphasia. A non-language area is still "wired" to the nearby areas, and any trauma may negatively affect the language areas by interrupting, mishandling, or cutting off communication betweent the language areas.[3] Just how interconnected the different parts of the brain are when language is involved has been up for debate for the greater part of the last century and a half, with such scholarly debates as those occurring in Berlin in 1874 which sought to answer whether the language centers could be defined and if damage to said regions would produce consistent results in most patients.[4]
Figure 1. Left hemisphere of a human adult brain. (source accessible to W&M students by clicking the image)

Damage to different language areas of the brain tend to produce different types of aphasia, but by no means is the resulting condition certain. General symptoms of aphasia from damage to the different areas are somewhat consistent, though combinations of symptoms are always possible due to the individual case.

Broca's aphasia is usually characterized by non-fluent, agrammatical speech, with major difficulty with spontaneous speech production.[5] Reading and auditory comprehension is maintained, but repeating and naming are often severely affected.[6]

Wernicke's aphasia is fluent but paraphasic, and audio comprehension is diminished.[7] Complicated grammatical structures are produced, but reading and writing ability may be situated at either extreme: severely hindered or completely unaffected.[8]

Global aphasia results from damage to both Broca's and Wernicke's areas, creating a disconnect among the transmitting language areas.[9] "Verbal stereotypes" such as swear-words, clichés, and idioms may be preserved, but otherwise syllables may be strung together to create meaningless "words."[10] It has been suggested that audio comprehension is especially good regarding concrete words and terms relevant to the individual sufferer.[11]

Conduction aphasia is characterized by fluency, and good comprehension both reading and auditory, but difficulties with repeating or echoing when prompted.[12] Those with conduction aphasia are aware of the errors they make in speech and thus attempt to correct their "phoneme substitution errors."[13]

Transcortical aphasia occurs in two forms: sensory and motor, wherein the former is characterized by poor comprehension and the latter by restriction of speech, though both forms have very good repeating/echoing ability.[14]

Anomia is an aphasic condition of "word-finding difficulties," and can be a somewhat positive improvement in recovery from Wernicke’s or conduction aphasia.[15]

Early studies of aphasia, up until the late nineteenth century, only addressed issues on a lexical level, such as the retrieval of words.[16] Through the early twentieth century, the "Wurzburg School" influenced the shift to sentence-level examination of aphasia symptoms.[17] A University of Berlin linguistics professor, Heymann Steinthal, created the term acataphasia to distinguish sentence-level difficulties within the different forms of aphasia.[18]

Moving Beyond the Word

Adolf Kussmaul (1822-1902), author of "Disturbance of Language: An Attempt in the Pathology of Language" (1877), was one of the first to suggest that some forms of aphasia extend past basic lexical problems into phrase structures and full-sentence syntax.[19] His school of thought descended from that of Steinthal, which was against the localizationalist model of language in the brain due to the limits it placed on language processes. Kussmaul criticized the medical approach of contemporary doctors and psychologists for the over-generalization of aphasia symptoms rather than differentiation between the processes affected.[20]

Figure 2. Adolf Kussmaul from the Clendening Library Portrait Collection.

In his 1877 work, he describes five disorders of aphasia, the last of which holds the most relevance to syntax:

  1. atactic aphasia
    • This is difficulty with the physical formation of words.

  2. amnestic aphasia
    • This refers to the difficulty retaining words as combinations of sounds.

  3. word deafness
    • This is a reduced capacity to comprehend words even though hearing is undamaged.

  4. paraphasia
    • This is a problem reconciling words and their concepts, so incorrect terms or unintelligible creations are used.

  5. agrammatism or acataphasia
    • This is the "inability to form words grammatically and to syntactically order them into a sentence."[21]

Kussmaul's linguistic approach sought to address the "impairments in grammatical formulation," and thus he subdivided his agrammatism into three forms.

  1. aphasic acataphasia, in which words are dropped which likely impact the conveyance of information in somewhat unpredictable ways, sometimes resulting in loss of content words.
  2. grammatical acataphasia, in which the words dropped are less important to the sentence structure, often described as telegraphic speech. The term "telegraphic speech" is often used as a substitute for agrammatism which is misleading because telegraphic speech is, in fact, grammatical.
  3. syntactic aphasia, which is now called paragrammatism.[22]

Agrammatism and Paragrammatism

Figure 3. Arnold Pick, from the "Founders of Neurology" series at the University of Illinois at Chicago.

Arnold Pick (1851-1924) continued the exploration of the "new psychology" developing in European medicine as it related to aphasiology.[23] He was a proponent of the sentence-schema approach, which theorized that one chooses specific words only after the word order is constructed in the mind.[24] Pick described only two forms of agrammatism: isolated agrammatism and pseudo-agrammatism. The former of these describes a "disorder of grammatical processing where the sentence schema is not available," which is an entirely syntactic persepective, observing that phrase fragments and common phrases tend to persist.[25] The latter refers to a perceived impairment, when the speaker prefers not to speak for some reason though thinking and syntax are unaffected;[26] this could perhaps describe situations in which speakers with speech impediments reduce their speaking habits, prompting some assessment of agrammatism.

Karl Kleist (1878-1960) defined paragrammatism as the disorder in which word order is maintained by phrases, clauses, and sentences but are strung together to create "confused sentence monsters."[27] He firmly believed that agrammatism and paragrammatism existed in "pure" forms only rarely, if at all, instead insisting that the two were almost always mixed to varying degrees.[28]

Henry Head (1861-1940) addressed the issues of syntactic aphasia (which is now synonymous with paragrammatism) as a "more or less severe disorder of rhythm and syntax. The patient speaks quickly, he speaks jargon, and prepositions, conjunctions, and articles tend to be omitted, polysyllabic words are slurred and pronounced badly."[29] His approach for the most part stood against the "diagram-makers" that illustrated the processes of speech as convoluted pathways, but he acknowledged the importance and holistic approaches of Kussmaul and Pick's descriptions. He provided a description of a pure form of syntactic aphasia, but like Pick, stated that such symptoms were unlikely ever found unmixed.[30]

Tesak and Code's compendium of aphasia scholarship offers a most comprehensive summary of the features of agrammatism: "...there may be a paucity or absence of function words in contrast to content words, omission of auxiliary verbs, impaired inflection of verbs, nominalization of verbs, impaired theta-role assignment from semantic and mapping from semantics to syntax, and tense-marking..."[31] These symptoms are not expected to be found packaged together in all patients, but symptom sets would include at least some of the above characteristics in speech. It bears mentioning that many documented cases that viewed aphasiology through the lens of syntax were completed in French and German, which has prompted the observation that agrammatism is even more apparent in inflected languages.[32] The authors of several cross-linguistic studies argue that agrammatism may be paragrammatism in English, suggesting the behavoir associated with dropped bound morphemes that still produce words is a substitution rather than the "error" associated with languages where dropping an inflectional ending does not create an actual word.[33]

Omission Trends

Agrammatism, due to its close ties to Broca's aphasia, involves the deletion of function words such as articles, conjunctions, prepositions, and auxiliary verbs. Chatterjee and Maher elaborate:
  • ...there are only a finite number of functors in any language; by contrast, the number of ‘open-class words,’ which are nouns, verbs, adjectives, and adverbs, is unlimited. Agrammatical patients also omit bound grammatical morphemes in their speech. Bound grammatical morphemes are parts of words that cannot stand by themselves and that modify the meaning of open-class words.[34]

In this manner, endings such as the "-ed" tense marker (preterit) may be omitted. Agrammatism is usually comprised of syntactic and morphological dysfunctions to varying degrees.[35] While these omissions may at first seem like simple lexical deletions for ease of articulation, the very nature of these words is heavily dependent on syntactic structures.

Further evidence of omission practice has been revealed by comprehension tests, such as those of Zurif, Caramazza, and Myerson's 1972 battery.[36] Both normal speakers and those with aphasia were presented with a brief sentence, and then provided cards with all combinations of three words from the original sentence. They were asked to choose the two words most closely related out of each set of three. Researchers compiled the answers and created structure trees. One example provided in Chatterjee and Maher is the sentence "The man was hurt." The normal speakers tended to pair the components of NPs together and components of VPs together: The + man, was + hurt. The aphasics produced man + hurt, which reveals comprehension of the idea of the original sentence, but with complete disregard for the function words.[37]

Mapping, Semantics, and "The Problem with Verbs"

The mapping hypothesis,[38] which states that thematic roles are "mapped" onto syntactic roles in language production, sheds some light on the inability of aphasics to comprehend or interpret reversible sentences. The ability to accurately map reversible sentences is aided by "lexical-semantic knowledge," that is, a basic understanding of the possible agents in question.[39] For example, in the following pair of sentences, the reversibility is limited by one's understanding of animacy:
  1. The girl kicked the ball.
  2. *The ball kicked the girl.
The girl is a living thing and serves as the agent in the first sentence. The second sentence is illogical because a ball is an inanimate object and therefore cannot perform the action of kicking. This seems obvious, but the judgment of such sentences is made possible by the underlying understanding of what a girl, a ball, and the act of kicking are. The thematic roles are therefore straightforward, and even in aphasics, the clues of prior lexical comprehension and experience allow determination of grammaticality and intelligibility. But by changing the inanimate object to a being with agency, the situation may be completely misunderstood by a person with aphasia.
  1. The girl kicked the boy.
  2. The boy kicked the girl.
The processing of the above pair of sentences is affected in most aphasics, resulting in confusion over which potential agent is the agent of the sentence, relegating the other potential agent to the role of patient. This misunderstanding is the basis for word order error, which is not denoted by haphazard choice of word placement. Illogical word order patterns are rarelyjust invented: Noun-Noun-Verb or Verb-Noun-Noun patterns are not created. [40] In general, Noun-Verb-Noun order is retained, but the first and second Nouns may be switched.

Saffran, Schwartz, and Marin (1980) conducted tests to specifically determine the role of animacy in agrammatical aphasics' sentence judgment. Their tests addressed both action and location, the first experiment involved the formation of sentences from pictures pairs.[41] The situations depicted involved two members, in three combinations of animacy: both animate, both inaminate, or one animate and the other inanimate. Agrammatical apashics performed better when the depictions involved differently animated things.[42]

A few studies of aphasics suggest that verbs are therefore the root of such problems. The key question becomes: Are verbs more difficult than nouns? To answer this, Lesser and Milroy point to the semantic differences between verbs that mean nearly the same thing, where theta-roles may create confusion regarding patient and theme, thus leading to unnatural verb choice.[43] Some aphasics have no trouble with verb-finding, but, as Lesser and Milroy explain: "For those patients who do have specific difficulties with verbs and the verb phrase, it seems that this may not be simply due to a greater inherent difficult which lies in verbs as such, but that the patients have specific difficulties in the operations which verb processing requires."[44]

Government and Binding Theory (Chomsky)

The Goverment and Binding Theory of Noam Chomsky has been applied to aphasiology in a several ways, providing a clearer picture of the levels of syntax affected in agrammatism. Deep structure is the form of a sentence where the grammatical features, and therefore semantic roles, are originally located. The surface structure is the sentence after movement. Following such movement, the surface structure will often contain a trace of some sort.

A 1988 study by Caplan and Hildebrandt[45] enumerated four categories of movement and their asociated traces in which agrammatical aphasics are likely to have difficulty:
  1. Wh-movement (questions and relative clauses which include wh-words)
  2. NP-raising (movement to a grammatical position, but no asigned grammatical role)
  3. other NPs due to "passivisation" and clefting
  4. PRO-gaps (ex: where pronouns are implied in CPs) [46]

Grodzinsky's 1989 study suggested the process of movement leads to omission: "...patients with asyntactic comprehension are unable to appreciate the grammatical link between the moved constituent and its underlying trace. They delete the trace in the S-structure of sentences and are unable to recover the deeper structure required to comprehend the meaning of the sentences."[47] The Trace-Deletion Hypothesis states that traces with theta roles are omitted or ignored when semantic roles are assigned.[48] Further, if the agrammatical aphasic must assign a theta-role, it is most often determined by the position in the sentence. [49]
Therefore, the concept of "referential dependency," described by Lesser and Milroy as overt, such as reflexive pronouns, or covert, such as empty traces,[50] becomes one the most difficult syntactic hurdles for aphasics.


  1. ^ Lesser and Milroy, 8, 13.
  2. ^ Lesser and Milroy, 13.
  3. ^ Reinvang, 25.
  4. ^ Tesak and Code, 75.
  5. ^ Lesser and Milroy, 8.
  6. ^ Reinvang, 25.
  7. ^ Reinvang, 27.
  8. ^ Reinvang, 27.
  9. ^ Reinvang, 28.
  10. ^ Reinvang, 28.
  11. ^ Reinvang, 28.
  12. ^ Lesser and Milroy, 8.
  13. ^ Reinvang, 29.
  14. ^ Lesser and Milroy, 8.
  15. ^ Lesser and Milroy, 8.
  16. ^ Tesak and Code, 73.
  17. ^ Tesak and Code, 123.
  18. ^ Tesak and Code, 74.
  19. ^ Tesak and Code, 85.
  20. ^ Tesak and Code, 85.
  21. ^ Kussmaul, 154-155, as cited in Tesak and Code, 85.
  22. ^ Tesak and Code, 85.
  23. ^ Tesak and Code, 126.
  24. ^ Tesak and Code, 128.
  25. ^ Tesak and Code, 128.
  26. ^ Tesak and Code, 128.
  27. ^ Kleist as cited in Tesak and Code, 129.
  28. ^ Tesak and Code, 129.
  29. ^ Tesak and Code, 135.
  30. ^ Tesak and Code, 135.
  31. ^ Tesak and Code, 212.
  32. ^ Menn and Obler's 1990 Cross-Linguistic Aphasia Study, as cited in Tesak and Code, 214.
  33. ^ Brief discussion of Menn, O'Connor, Obler, and Holland's 1995 cross-linguistic study in Chatterjee and Maher, 142.
  34. ^ Chatterjee and Maher, 133.
  35. ^ Chatterjee and Maher, 134.
  36. ^ As summarized in Chatterjee and Maher, 139.
  37. ^ Chatterjee and Maher, 139.
  38. ^ Lesser and Milroy, 88.
  39. ^ Chatterjee and Maher, 137.
  40. ^ Chatterjee and Maher, 137.
  41. ^ Saffran, Schwartz, and Marin, 264.
  42. ^ Saffran, Schwartz, and Marin, 268.
  43. ^ Lesser and Milroy, 87.
  44. ^ Lesser and Milroy, 88.
  45. ^ As cited in Lesser and Milroy, 92-93.
  46. ^ Lesser and Milroy, 93.
  47. ^ Grodzinsky as cited in Chatterjee and Maher, 140.
  48. ^ Beretta et al. 727-728, after Grodzinsky 1994.
  49. ^ Beretta et al. 727-728, after Grodzinsky 1994.
  50. ^ Lesser and Milroy, 93.

Works Cited

Beretta, Alan, Carolyn Harford, Janet Patterson, and Maria Pinango. "The Derivation of Postverbal Subjects: Evidence from Agrammatic Aphasia." Natural Language & Linguistic Theory 14, no. 4 (Nov. 1996): pp. 725-748.

Chatterjee, Anjan, and Lynn Maher. "Grammar and Agrammatism." In Aphasia and Language: Theory to Practice, edited by Stephen E. Nadeau, Leslie J. Gonzalez Rothi and Bruce Crosson. New York: The Guilford Press, 2000.

Lesser, Ruth, and Lesley Milroy. Linguistics and Aphasia: Psycholinguistic and Pragmatic Aspects of Intervention. London: Longman, 1993.

Reinvang, Ivar. Aphasia and Brain Organization. New York: Plenum Press, 1985.

Saffran, Eleanor M., Myrna F. Schwartz, and Oscar S. M. Marin. "The Word Order Problem in Agrammatism: Production." Brain and Language, no. 10 (1980): 263-280.

Tesak, Juergen, and Chris Code. Milestones in the History of Aphasia Theories and Protagonists. Hove, UK: Psychology Press, 2008.