Deb makes two very important points in her introductory article to the special issue. I'd like to highlight these for discussion.

Her first point is that "despite the long history of scholarly work on language and schizophrenia, mainstream psycholinguistic and neurolinguistic researchers are often unaware of this literature." This is so true. Many of my colleagues in psycholinguistics seem truly interested and excited when they hear how their work might be applicable to understanding language and communication disorders in schizophrenia, but tell me that they have been unaware of its relevance. I think that the reasons for this are largely historical. There is a rich tradition of studying language disturbances in patients with discrete brain lesions; indeed, many of the older neuropsychological models of language stemmed directly from the study of the aphasias. Also, as a neuropsychiatric disorder that presents in adults, schizophrenia contrasts with developmental communication disorders like dyslexia and autism, which both receive the attention of speech and language...  Read more

Deb makes two very important points in her introductory article to the special issue. I'd like to highlight these for discussion.

Her first point is that "despite the long history of scholarly work on language and schizophrenia, mainstream psycholinguistic and neurolinguistic researchers are often unaware of this literature." This is so true. Many of my colleagues in psycholinguistics seem truly interested and excited when they hear how their work might be applicable to understanding language and communication disorders in schizophrenia, but tell me that they have been unaware of its relevance. I think that the reasons for this are largely historical. There is a rich tradition of studying language disturbances in patients with discrete brain lesions; indeed, many of the older neuropsychological models of language stemmed directly from the study of the aphasias. Also, as a neuropsychiatric disorder that presents in adults, schizophrenia contrasts with developmental communication disorders like dyslexia and autism, which both receive the attention of speech and language therapists in childhood.

This disconnect between the fields of psycho/neurolinguistics and the psychopathology of schizophrenia has had two important consequences: (a) we have not been drawing enough from models of normal processing in formulating our hypotheses about abnormal language processing and production in schizophrenia (we can perhaps consider this issue in part 1 of our discussion); (b) there exist very few cognitive remediation approaches targeting language and communication in patients with schizophrenia (important for discussion in part 3).

Deb's second important point is that "schizophrenia, like cognitive neuroscience itself, forces us to abandon modular psycholinguistic approaches that have historically relegated real-world contextualized language understanding to ‘‘pragmatics’’ or ‘‘problem-solving.’’ Again, so true, and hopefully something that we can consider in part 2 of our discussion. Deb is right in that modular approaches have been challenged in psycho- and neurolinguistics. Still, there remains a general idea that the language system is insulated from more general cognitive functions (executive, working memory, attention). Perhaps this again stems from the traditional neuropsychological approach which examined language disorders in patients with discrete lesions. One cannot study language in schizophrenia without fully taking on board our patients' other types of cognitive deficits and without considering the fact that they do not have discrete lesions but rather multiple, widely distributed structural and functional abnormalities.

On the other hand, I do think that we shouldn't be content with conceptualizing language dysfunction as being simply reducible to working memory or executive dysfunction plus semantic memory problems. We need to be able to define the interactions between these cognitive processes within normal behavioral and cognitive neuroscience models of language processing and production.

One of the thematic aspects that recurs in considering language dysfunction involves critical aspects of neurodevelopmental considerations. Ruth Condray (2005) suggested that a number of developmental communalities were likely to contribute to both schizophrenia and dyslexia. However, the concept of shared linguistic substrates contributing to non-psychopathological disorders seems to have been resisted by the developmental linguistic community to date, although Pasko Rakic (e.g., Rakic, 2009) has suggested that aberrant cortical organization during development may underlie a number of conditions including schizophrenia, dyslexia, and autism. It is interesting that in the past few years, converging genetic contributions to these disorders have begun to be found.

References:

Condray, R. Language disorder in schizophrenia as a developmental learning disorder. Schizophrenia Research, 73:5-20, 2005. Abstract

Rakic P. Evolution of the neocortex: a perspective from developmental biology. Nat Rev Neurosci. 2009 Oct;10(10):724-35. Review. Abstract

View all comments by Stuart Steinhauer

I outline the four-quadrant theory as follows:

1. The human brain differs from that of other primates by the addition of a "torque" from L frontal to R occipital as described by Gratiolet in 1840.

2. This renders the human brain four-chambered in terms of areas of association cortex by comparison with the two chambers (anterior and posterior) of the generalized primate brain.

3. This gives directionality to inter-hemispheric connections: R > L anterior, L > R posterior, to generate the "language circuit."

4. Within each quadrant there is a characteristic type of engram: L posterior—speech perception > R posterior—meaning > R anterior—thought > L anterior—speech production.

5. These can be understood as the bidirectional connections postulated by de Saussure of a signifier (word patterns—L) with its signifieds (meanings or thoughts—R), or the four compartments of language envisaged by Chomsky (articulatory- perceptual on the L and conceptual-intentional on the R).

6. Nuclear symptoms reflect the breakdown ("leakage")...  Read more

I outline the four-quadrant theory as follows:

1. The human brain differs from that of other primates by the addition of a "torque" from L frontal to R occipital as described by Gratiolet in 1840.

2. This renders the human brain four-chambered in terms of areas of association cortex by comparison with the two chambers (anterior and posterior) of the generalized primate brain.

3. This gives directionality to inter-hemispheric connections: R > L anterior, L > R posterior, to generate the "language circuit."

4. Within each quadrant there is a characteristic type of engram: L posterior—speech perception > R posterior—meaning > R anterior—thought > L anterior—speech production.

5. These can be understood as the bidirectional connections postulated by de Saussure of a signifier (word patterns—L) with its signifieds (meanings or thoughts—R), or the four compartments of language envisaged by Chomsky (articulatory- perceptual on the L and conceptual-intentional on the R).

6. Nuclear symptoms reflect the breakdown ("leakage") of the insulation of the four compartments: thought insertion, withdrawal from posterior to anterior; third-person voices, and thoughts spoken aloud from anterior to posterior.

7. These and other symptoms represent a "failure of lateralization" of the components of language.

8. This failure constitutes a component of the diversity associated with the emergence of the torque.

The genetic basis lies not in "polygenes," for which there is no evidence, but in the evolutionary history of the Protocadherin11XY gene pair that characterizes the hominin lineage including the "sapiens" event that represents the transition to modern Homo sapiens.

References:

Crow TJ. The "big bang" theory of the origins of psychosis and the faculty of language. SCHRES 2008;102: 31-52. Abstract

Crow TJ. Who forgot Paul Broca? J Linguistics 2005; 41: 133-156.

Crow TJ Schneider's nuclear symptoms as the key to the structure of language. Neurol Psychiat Brain Res 2007; 14: 87-94.

Williams NA et al. Accelerated evolution of the Protocadherin11XY gene pair. AMER J MED GENET 2006 141B: 623-633. Abstract

An important ongoing endeavor is to delineate the link between neurophysiological abnormalities on the one hand, and the clinical symptoms and functional impairment seen in schizophrenia (e.g., the social skills deficits examined by Mueser et al. in this issue). The report by Borofsky et al. contributes to this effort by reporting a link between decreased left inferior frontal gyrus and superior temporal gyrus activation and disorganized speech in childhood-onset schizophrenia.

This endeavor will potentially be assisted by more sensitive quantitative measures of language production disturbance (as demonstrated by Elvevag et al.), which draw on current methods from computational linguistics such as...  Read more

An important ongoing endeavor is to delineate the link between neurophysiological abnormalities on the one hand, and the clinical symptoms and functional impairment seen in schizophrenia (e.g., the social skills deficits examined by Mueser et al. in this issue). The report by Borofsky et al. contributes to this effort by reporting a link between decreased left inferior frontal gyrus and superior temporal gyrus activation and disorganized speech in childhood-onset schizophrenia.

This endeavor will potentially be assisted by more sensitive quantitative measures of language production disturbance (as demonstrated by Elvevag et al.), which draw on current methods from computational linguistics such as latent semantic analysis.

These new methods, along with the progress made in characterizing the often subtle and complex language processing abnormalities in schizophrenia (e.g., the study by Salisbury, and review by Ditman and Kuperberg) hold promise for developing more standardized neurophysiological and behavioral assessments for clinical applications (e.g., diagnostic/prognostic indices, or biomarkers for response to pharmacological treatments). One challenge in standardizing such instruments is the different versions and norms that would be required for populations with different linguistic experience.

First, I would like to thank Deb Titone for organizing this interesting discussion. Second, I’d like to follow up on Gina Kuperberg’s encouragement not to limit our thinking about language dysfunction to the problems in working memory, executive dysfunction, and semantic memory that are clearly associated with schizophrenia. Her point is a good one, and I think it bears on an assumption that is driving much of our recent work: that cognition links with clinical symptomatology in straightforward ways. However, the data don’t always support our predictions regarding which aspect of language dysfunction will be correlated with which specific clinical symptoms, and in which direction. As an example, Michael Kiang notes in his   Read more

First, I would like to thank Deb Titone for organizing this interesting discussion. Second, I’d like to follow up on Gina Kuperberg’s encouragement not to limit our thinking about language dysfunction to the problems in working memory, executive dysfunction, and semantic memory that are clearly associated with schizophrenia. Her point is a good one, and I think it bears on an assumption that is driving much of our recent work: that cognition links with clinical symptomatology in straightforward ways. However, the data don’t always support our predictions regarding which aspect of language dysfunction will be correlated with which specific clinical symptoms, and in which direction. As an example, Michael Kiang notes in his excellent review of the odd speech symptom in schizotypy that the data appear to reflect increased activation of weakly related ideas at shorter SOAs, and underactivation of both weakly and strongly related ideas at longer SOAs. Many of us, I think, would predict that these activation patterns would be related to clinical thought disturbance in intuitively straightforward ways. In the patient data from our Pittsburgh lab (Dr. Stuart Steinhauer, Director), we have observed the expected pattern at a long SOA (that is, better language response was correlated with reduced clinical thought disturbance). However, no relationship was observed between language response at the short SOA and clinical thought disorder. Moreover, language response at the short SOA was correlated with symptoms of paranoia, and in an unexpected direction: more typical or normal language response was associated with an increase in paranoia symptoms.

All of the above seems to suggest that we need to think carefully about how our neuroscience models of cognitive disorder and language dysfunction in schizophrenia can direct our understanding of its hallmark clinical symptoms.

References:

Kiang M. Schizotypy and language: A review. Journal of Neurolinguistics 23;3 May 2010:193-203. Abstract