Thought-provoking question.  Fundamentally, it depends on one's assumptions regarding the definition of language, right?  Cognitive linguists, for instance, assume a relationship between language and cognition in their definition of language.  In contrast, structural linguists assume that linguistic capacity and cognitive capacity are separable.  Yet another cohort of linguists, functional linguists, view language as a set of communicative "tools in a toolbox" that  function in context to convey meaning, supported by underlying cognition.  Among clinical/research aphasiologists, I suppose we can claim representative adherents within each of these camps, some of whom adhere more tightly to one set of assumptions, and others who adhere less tightly to just one set of assumptions and who act as bridges between the various perspectives.  One's set of assumptions holds implications for both clinical practice and modeling. For example, currently influential aphasia typologies are structural in their orientation, and the field of aphasiology could be revolutionized if alternative/parallel typologies based on cognitivist or functionalist perspectives were to be proposed. 

The aim of the Fluency Profiling System (Hird,  Kirsner, Oehmen,Little &Dunn) is to move away from structuralist assumptions and associated typologies. Current neuroscience supports a distributed network understanding of processing. A conceptualization of Aphasia that is inclusive of all aspects of cognition and linguistics is more in keeping with what is now known about the supporting neural mechanisms. A new way of classifying the functional consequences of disruptions to that network is therefore required. Verbal output is surely a product of all underlying processes involving motor, cognition and linguistics. The high variability that is observed in the characteristics of verbal output following neural impairment across contexts ,over time and in response to a variety of treatment strategies reveals the insufficiency of the current classification systems describing  "aphasia".

 I'm with you Chris! We are currently working on a new classifications that captures these concepts. Our limitation is in obtaining  enough good quality recorded samples to analyse through our automated system. We are currently processing Aphasia Bank samples.

Thanks for your responses Gloria & Kathryn.

Right Gloria! Yes, of course how we see 'language' influences how we see aphasia. A functionalist perspective, and related perspectives, is much better placed getting to grips with the problems of an aphasic person trying to cope with communicating in a communicating world than a cognitive linguistic perspective. How language functions for us as social & communicating beings looks very different to 'language' swirling around some cognitive space.

Your new work looks exciting Kathryn. But a  new 'classification'? Your system sounds to dynamic & fluid to produce such a thing. Maybe yet another classification is something to avoid! I suppose too that 'all aspects of cognition and linguistics' will change, and grow, as neuro-cognitive research reveals yet more associations & dissociations?

The Fluency system is actually producing some clear and objective results where individuals can be plotted  against a normal distribution for a number of different parameters. We use an EM algorithm and start with the  raw acoustic signal thereby avoiding any a priori assumptions based on other classifications systems.  What we can actually see are patterns of performance that can be correlated with fMRI data and perceptual judgement to ensure face validity. The longer term aim is not to produce yet another classification system, rather to describe individual performance over time and then to identify which parameters predict that performance.

I hypothesise that  an interaction between neurophysiological markers (genetic, co-morbidities) cognitive (processing capacity,  attention,  new learning etc) and environmental (motivation, care burden ) rather than site and volume of lesion will be the best at predicting the trajectory of function post lesion (of any etiology). 

The Fluency system is very sensitive and is great for quantifying change over time. The data we have so far - approx 100 cases with a boston Dx of Brocas aphasia shows evident patterns that map onto gross perceptual judgments but also quantify the individual differences with a high granularity.

So I don't think its wishful thinking to produce an alternative way of thinking about "aphasia"!

I just downloaded your paper Kathryn, and look forward to reading it. I welcome any new approach to aphasia, especially ones that circumvent (neo)classical perspectives. I suppose a question that immediately occurs is, how does the system deal with non-expressive aspects of aphasic impairment, like comprehension?

Very interesting question! The answer depends, I guess, on whether an ability is defined by what it is an ability to do – quite conventional – or by theoretical constructs inferred to sustain this ability. ‘Aphasia’ is an impaired ability to use language for interpersonal communication. It has concrete behavioural manifestations in everyday life and in clinical assessment. So aphasia is a language disorder. Language can be shown to depend on a set of subordinate abilities, some of which are non-observable constructs and probably themselves further divisible into more fine-grained functions. However, it does not follow from this that aphasia is not a language disorder. Perhaps these subordinate abilities can be plausibly considered constitutive of normal language use, rather than eliminating language and aphasia by reducing it to its ‘horizontal’ functions?

It is indeed a very interesting question. I agree with Gloria who suggests that the answer is often a matter of emphasis depending on the theoretical framework we adhere to. It seems that there is a great deal of aphasic data supporting both approaches (i.e. inclusive and exclusive of cognitive aspects of speech performance). Within the functionalist approach, I think the usage-based perspective, which sees the linguistic structure as emergent from domain-general cognitive processes, may be a good candidate to account for the cognitive and neural reality of linguistic performance without denying the regular patterning across languages. Thus, my answer to the question would be that aphasia is a linguistic disorder, since it affects the specific aspects of the linguistic structure captured at a particular stage of its development.

Thanks Per and Anna for responding. I understand what you say - clearly my question is a 'neurocognitive' one. Of course, for an aphasic person and their family, and the people working with that person, there is a devastatingly obvious communication disability. I think my question asks if that devastatingly obvious communication disability is 'caused' by damage to processes that underlie the processing of what we call language.

How we approach an answer has implications for our treatment and intervention.

I understand quite well your somewhat provocative proposal. In fact, to a certain extent, it is not that much provocative if we take into consideration the fact that structural complexity as viewed by linguists (whatever might be their favorite model) often goes along with processing complexity at the psycholinguistic, cognitive, level.

Letme take two examples : (i) if embeddeed relative clauses tend to be more difficult to procès (in children, in aphasics…) is it for striuctural reasons or for processing reasons. This type of relative clause clearly is more complex to handle for mnesic reasons (we have to keep in mind the beginning of the main clause, while listening to the relative, before processing the end of the main clause. (ii) is anaphor résolution a linguistico problem or a mnesic (WM) problem ?

From a neurocognitive point of view, there is another issue which deserves attention.

In the human brain, language came to develop itself in parts of the cortex which, in non-human primates, were processing « action ». Fairly recent papers tend to show that action processing and action verbs processing maybe disturbed together (Luria provided nice examples of that tendency in patients with frontal lesions). Brain Imaging, in normals, tend to show the same tendency. Such a kind of data, if confirmed, might be an indication that underlying an apparently « linguistic » deficit, might exist a deeper cognitive processing problem.

Of course, counter-examples might also be looked for : if, for instance, we think that an apparently suntactic problem might derive from a deeper WM problem (in both production and compréhension), do WM problems exist, in other areas of cognition, without disrupting language processing. Here we go again : looking for « dissociations »…

I'd like to play devil's advocate, and suggest that many of those apparently general capabilities that support language are in fact language!

For example, there is much neuropsychological and lesion evidence to suggest that working memory for language is supported by the very same systems that represent the relevant language elements (in particular, phonological representations). The same goes for control. As we discover more about the control functions of the prefrontal cortex, we find many aspects of control that are domain specific - for example, the left inferior gyrus seems to be important for control of language specifically. Plus, we know from DTI studies that this region has rich interconnections with  posterior language regions. Finally, we know that speech apraxia can occur in the absence of other praxic disturbances, So it really is praxis for language specifically.

So to me, aphasia is a disturbance to one or more cognitive processes that are essential - and relatively specific - to aphasia. They would not include processes that are genuinely non-language specific (for example patients with anterior cingulate lesions don't say much, but they do much at all of anything, so I wouldn't count that as language). But anything process that, when impaired, has a disproportionate impact on language, is - to my way of thinking - language!

Very interesting question indeed. As Dr. Nespoulous noted, we must see language within an evolutionary context. Comparative anatomy has shown that there are several homologous cortical areas, as well as white matter interconnections between humans and other primates. Humans, due to several pre-adaptations thet emerged at some point of the evolutionary timeline (such as the L-shaped tract) became able to articulate speech.

Moreover, several studies have shown that aphasic patients demonstrate various deficits, that are not language-specific (e.g. impaired visuospatial STM/WM). Brain imaging studies are in accordance with these findings, since they point to a perisylvian network supporting STM/WM.

From an evolutionary perspective, it is highly plausible that language functions emerged via natural selection, based on physiological pre-adaptations, such as WM cortical networks consisting of interconnecting frontal, temporal, and parietal areas.

In this sense, aphasia is a cognitive rather than linguistic disturbance.

Thanks for your response Dimitrios. I seem to have 'lost' the answers of Jean-Luc and Carolyn. Does anyone know how to retrieve them?

Of course, Dimitrios, language & speech evolved, and probably from pre-existing abilities. That language evolved from pre-existing action skills is the currently most popular suggestion from people like Corbalis and Arbib: the close relationship between speech & gesture accompanying speech would appear to be more than a coincidence. But I don't think that necessarily means that certain language processes, like perhaps certain phonological processing skills, are not, as Carolyn suggests separate and independent of other basic functions. Even if they are 'independent', however, phonological knowledge, like any stored knowledge, needs storing in memory.

Broca's area is more than the area damaged in many cases of Broca's aphasia - Hagoort (2005) identified an enlarged area he calls 'Broca’s complex'. Activation of this region is seen while experimental participants engage in semantic (BA 47 and 45), syntactic (BA 45, 44, 46) and phonological processing (BA 45, 6) with substantial overlap. Hagoort suggests it has a cognitive selection and binding or unification role for information retrieved from the mental lexicon  and emphasises its role in maintaining information on-line while binding operations can take place. Thompson-Schill (2005) stresses the importance of Broca’s area in the selection of information from competing sources. As Jean-Luc said, it is also engaged in the processing of actions.

Language was defined in my schooling as THE INTENTIONAL USE OF SHARED ABSTRACT SYMBOLS.  Aphasia was defined as a LOSS OF LANGUAGE SKILLS and so ...

Yes,  aphasia is language disorder.  If my instructiors were correct, It' really a matter of definition of the term,  then the real questions are whether memory and praxis and all the other lateral skills are necessarily part of language. 

Language skills are demonstrated with a child's first word attempts and progress to include semantic and syntactic constructs.  Audition, eye gaze, postural motion are involved with the neurodevelopmental progression starting neonatally are all necessary for neurological development of the childs language skills.  The inclusioin of visual langauge (reading & writing) creates a significant jump in cognitive abilites, evident in the contrast in the cognitive abilites between literate and illiterated groups of people. 

Neurologically language(ie:abstract information) processing goes dorsal -ventral, laterally, and superior-inferior in the brain as the information is coordinated.  Auditory-Visual processing and executive function coordinates experience with the information (this is dorsal-ventral). Localization for spacial awareness and processiong, prosody-melody and meter, gist-detail coordination are lateral across the corpus callosum-between the hemispheres. Feed forward and feed back for praxis and inititiation are superior-inferior through the brain and cranial nerves. Remember that  sign language is a language too. Some may question if interpretive dance is a use of shared symbology. Wrestle with that sometime! 

Memory is a matter of latent imaging being processed, integrated, and reprocessed; to reiforce previous engrams that would othewise degrade; increase longer retention; and eventual learning for longterm memory.  Where this involves abstraction it should be considered language by definition.

Motion is not necessarily based on abstraction.  More oftern than not it is indemic to structure and function.  When the motion is programmed as part of the abstracted intent then we include it as part of language. Articulation, phonology, prosody, semantic marking.

Chris, I have become more interested in this question since you posted it. In acquired language disorders, we assume a fundamental dichotomy between "Language" and "Non-language" skills. We assume non-language impairments can affect language, but these shouldn’t be considered as language. At some level, this seems justifiable - if you can't speak because you can't move your articulatory musculature, that seems clearly to be a non-language impairment affecting your language abilities – not a language disorder per se.

But this dichotomy becomes increasingly difficult to justify as you move up to higher cognitive skills. For example, it has been suggested from time to time that some individuals may have a difficulty understanding sentences not because they have a language disorder, but because they have a problem with working memory. Given what we now know about the reliance of verbal working memory on language processing abilities, this suggestion no longer makes a lot of sense. These two sets of skills are two sides of the same coin.

Things get even more muddy when we consider the recent research showing that individuals with nonfluent aphasia fail on certain “frontal” tasks like the Stroop task, but not with the same language materials on more standard reading/naming tasks (Hamilton & Martin, 2005, 2007; Scott & Wilshire, 2010). This is a little reminiscent of the selection deficits reported by Thompson-Schill and colleagues that you mentioned. Is this a language problem, or not? Randi Martin and myself believe that this is indeed a language problem, and our reasoning is that: a) the same difficulties are not observed in domains other than language (therefore, this is a language-specific skill); and b) the abnormality we are measuring in this task contributes significantly to the language problems of these individuals in naturalistic settings. That is, this difficulty managing or controlling the activation of language representations – which is so clearly seen in the Stroop situation - is one of the fundamental problems underlying anterior, Broca’s type aphasia.

In sum, I think we need to move away form the idea of opposing language and cognition, and situate language within cognition. The question should not be “Is it language?”, but rather, “Is it language-specific?” (does it affect activities that don't involve language?) and “Is it critical for effective language functioning”?

Carolyn

Thanks for your continued interest Carolyn. I don't think I disagree with anything you say, at least not seriously.

Maybe I can turn the question towards 'symptoms' for another perspective. I think there is good evidence that a number of classic symptoms of aphasia - characteristics by which we recognise someone has aphasia, and can indeed 'diagnose' aphasia -  may be more to do with other cognitive processes that language per se. Jargonaphasia has been characterised by a number of people in the past, and more recently, as a symptom that does not is 'nonlinguistic', although having 'linguistic' features. There is a lot more to neologistic jargonaphasia than a language problem and the underlying cause of jargonaphasias appears not to be linguistic, but a denial resulting in a mixture of anosognosia, delusion and euphoria, and even paranoia. The individual appears unable - not just unwilling - to understand they are ill because of damage to crucial areas in the parietal lobe. The jargon produced by the individual appears to be an anosognosic adaptation to aphasia, which appears to be the primary symptom. Far from believing there is anything wrong with his/her speech, the speaker may feel especially eloquent and precise (Weinstein & Puig-Antich, 1974; Rochford, 1974; Nespoulous, Code, et al, 1998). Comprehension can often be good enough to recognise jargon when they hear it from another resulting from some impairment in self-monitoring. As you know, the jargon is made up of formulaic, clichéd, humorous, irrelevant utterances, produced fluently and verbosely, but has little representational content, and the jargon itself appears to reflect phonological and semantic access problems, but these access problems do not appear to be the cause.

Another classic symptom, the recurring utterance or speech automatism of the Big Broca's, or global, aphasic person, identified originally in Leborgne (/tan, tan/ and Sacré nom de Dieu), and very commonly observed in such people, appears to be a result of the failure of those frontal processes you discuss. These utterances appear to be the result of failed inhibition or attention, and seem unrelated to linguistic problems directly, although they can be described in linguistic terms of course  (Alajouanine, T. (1956) Verbal realization in aphasia. Brain, 79, 1-28.. Code et al. (2009). Opportunities to say ‘yes’: rare speech automatisms in a case of progressive nonfluent aphasia and apraxia. Neurocase, 15, 445-458. These utterances are 'positive' symptoms in the Jacksonian sense.

Of course, non of this detracts from your points on whether syntactic comprehension and other impairments are due to WM impairments or auditory-verbal STM problems, that are language-specific impairments, or other language-specific impairments related to some frontal-executive problems. 

So there may be some grounds for suggesting language-specific and non language-specific features of aphasia?

Chris, your examples are really interesting ones. But having reflected on them, it still seems to me that we don't need to go beyond even the conventional definition of language function to come up with a really elegant explanation of them. The anosognosia you describe in jargonaphasia - I've never personally seen it without accompanying comprehension deficits. So an account in terms of accompanying comprehension does seem to me to be the most parsimonious. Sure, maybe some individuals may be better at telling others' jargon from their own, but that's a much less demanding task. You simply have to listen to the speech, you don't have to do that online at the same time as formulating your own speech plan. Plus, it might be easier to detect errors involving words/structures that one doesn't have a problem with personally - and these are likely to be more common in the speech of another person. Perhaps I should look over your paper in more detail to see if there’s more there that challenges this view, but from what you say, I’m still skeptical.

Similar arguments can be made for automatisms. If you think of language production in the context of an incremental learning type framework like Dell, Oppenheim et al (2010, LCP, I think), then utterances that are available to a person when none others are – for example, in the acute phase - will most likely be repeated often, and become further favoured (and alternatives further disadvantaged) through the accompanying incremental weight changes. The speech intention, if you like, gets “channeled” into these well worn, overlearned paths. Even the fact that these repeated utterances are often expletives doesn’t seem that surprising to me. The powerful emotional valence of some expletives might make them particularly easy to access when no other words are. Then the repetition of the error further increases the bias over time.

I'm not saying people with aphasia have only language problems in the conventional sense - that everything other than their ability to produce and/or understand language is just fine and dandy. In fact, quite the opposite, I'm saying that much of what we conventionally consider to be outside of language (reasoning, monitoring our language behaviour and controlling our thoughts and actions in the verbal domain) are emergent properties of our ability to use language. So language impairment - as conventionlly defined - affects these too.

I will take a look at those papers, though. Interesting issues to reflect on!

Chris,

There is a new book published relating to this by Dr. Martin Albert and Dr. Dalia Cahana-Amitay if you would like more about this topic.

Thank you Paula. I know of the book and hope to see it soon.

It is important to note that aphasia in childhood is quite different to aphasia in adulthood, since failure to recognize this results in totally inappropriate adult neurological models being applied to children.  Whilst the adult model is relevant to cases of aphasia from known cerebral trauma in young children, the condition of developmental aphasia still attracts a lot of coverage on Google.   Congenital aphasia, severe speech/language delay not traceable to any known brain insult was nevertheless generally deemed and widely accepted to be of primary cortical origin. The first challenge to this was when the Ewings performed careful audiometry on such typical cases, finding many examples of selective high tone hearing loss.  Then Vygotsky & Luria established the fundamental principle that  the most damaging lesions in infancy were lower peripheral ones, for adults it was the higher cortical ones.  Thus Luria pointed out that even "minor" degrees of hearing loss from otitis media in infancy could lead to aphasia.   Even so, I spent the first part of my career in audiology doing careful audiometry on a string of referrals from a specialist school for aphasics, in all of whom the deafness had been missed.

So, to come to a very simple conclusion: developmental aphasia is neither a language disorder nor a cognitive disorder, but, at least in the first instance, a sensory perceptual disorder, though not necessarily a permanent one.   If a peripheral auditory disorder has not been found, it is because people have not looked hard enough, or at the right time, or because they suffer from specific imagination disorder themselves.  After all, many, but not all, cases of known congenital deafness due to proven cochlear damage, where any cortical lesions have been excluded, have severe speech and language problems clearly secondary to their deafness.

Dr Gordon,

Thank you for your interesting response. Developmental aphasia is something I know little about, so your summary is particularly interesting and useful.

The nature of aphasia is unlikely to be elucidated by linguists or philosophers without reference to clinical observations on the development of speech and language and its dissolution, after strokes for example.    HB Wilbur has a still relevant discussion of these issues (Am J Insanity 1867;24:1):

Comments

Thanks for your interesting response, Dr Gordon. I'll have to check out Wilbur.