Comparative aphasiology is the systematic comparison of aphasia symptoms across languages, including the comparison of acquired reading problems across languages and writing systems. The goal of study is to ensure that theories that claim to account for the various constellations of aphasic symptoms can handle the similarities and differences seen in aphasic speakers of languages of different types (see Menn, 2001, for a discussion of language typology in the context of comparative aphasiology). Serious experimental and clinical comparative work began in the 1980s; however, researchers long before that time understood that comparative data are essential. Otherwise, general theories of aphasia would depend on data from the few, closely related languages of the countries in Europe and North America where research in neurolinguistics was being undertaken: English, French, German, and, for a time, Russian.

The clearest example of such a premature class of theories was the “least pronunciation effort” approach to agrammatism. This approach focused on the prevalence of bare-stem forms in agrammatic output in English, French, and German and suggested that such forms were used to avoid extra articulatory effort. A more phonologically sophisticated theory (Kean, 1979), relying on the same small database, proposed that agrammatic speakers were constrained to produce only the minimum phonological word. However, the languages then available for study have bare-stem forms that are not only short but also grammatically unmarked (mostly first-or third-person singular present tense of verbs and the singular of nouns) and very frequent. Therefore, the data could also support a markedness-based or a frequency-based account. Or the underlying problem could be morphological or morphosyntactic instead of articulatory or phonological.

A reason to think that the problem is actually morphological (problems with retrieving inflections) or morphosyntactic (problems with computing which inflections the syntax demands) is that endings of participles and infinitive forms are better preserved than other verb forms in English, Italian, French, and German. Is this because they have no tense marking or no person marking? Or is there another reason?

To test the hypothesis that a problem with inflections is a specific type of problem with morphosyntax, we look for languages where inflections are controlled in different ways. For example, if problems with verbs are blamed on difficulty with person or number agreement, we should see whether verb problems also exist in languages without agreement, such as Japanese. Or the problem may be deeper than morphosyntax: the verb problem could be due to a semantic difference between nominal and verbal types of elements, as suggested by experimental work on Chinese, which has no agreement or inflection (Bates, Chen, et al., 1991; Chen and Bates, 1998). Or perhaps multiple factors interact—a more difficult claim to test.

Another type of problem that demands a comparative approach is the question of why there are so few adjectives in aphasic language. From, say, a German-centered point of view, we might ask, Is this a conceptual problem, an agreement problem (nonexistent in English, so that could not be the sole problem), or a problem in inserting elements between article and noun (in which case, Romance languages, where most NPs have article-noun-adjective order, should not show the effect)?

Issue after issue requiring a comparative approach can be listed in the same way. Would relative clauses that do not require movement (as in Chinese or Japanese) be deployed better than ones that do? Is the observed problem with the passive voice in English to be explained in terms of movement rules and traces, in terms of its morphological complexity, in terms of its low frequency, or in terms of its pragmatic unnaturalness in a single-sentence test paradigm? Are irregular verbs preserved better than regular ones because of their generally greater frequency or because, as some theorists claim, they are stored in different places in the brain or deployed using different mechanisms? (Jaeger et al., 1996; Pinker, 1999).