Sunday, March 8, 2009

The Battle for Broca's area

David commented previously on Grodzinsky and Santi's (2008) TICS paper on the function of Broca's area. Justifiably in my view, David expressed his disappointment that an obviously multi-functional region was being linked to a single function, syntactic movement. In addition, a recent reply to the paper in TICS by Roel Willems and Peter Hagoort (2009) highlighted the possible role of Broca's area in semantic processing, an issue that was not addressed at all in the Grodzinsky and Santi paper. So there is no shortage of critics of the opinion piece. I might as well join the party...

By way of reminder, Grodzinsky and Santi discuss four main hypotheses regarding the function of Broca's area: action comprehension, working memory, syntactic complexity, and syntactic movement. They argue for the last as being the one that is best supported by the data. Having dabbled in theoretical explanations of the pattern of comprehension deficits in Broca's aphasia, I'm sympathetic to the syntactic movement account. However, there are two problems with the proposed link to Broca's area (BA 44/45). One is that the argument against the working memory theory is exceptionally weak and the other is that the comprehension deficit does not appear to be linked specifically with Broca's area.

The working memory theory. The idea behind the working memory theory is that sentences containing syntactic movement require additional working memory resources to process and Broca's region (or more accurately, the lesions that produce Broca's aphasia and agrammatic comprehension) is(are) critical for working memory. G&S argue against this position on the basis of "preliminary studies" (p. 477) of the comprehension of reflexive constructions, such as Mama Bear touched herself, which they suggest requires working memory. Six Broca's aphasics were able to comprehend such sentences "contrary to the prediction of a WM deficit account" (p. 477). Well, what if the amount of working memory required to comprehend Mama Bear touched herself is less than the amount of working memory required to comprehend The cat that the dog chased was very big? If there are working memory differences between these constructions, which seems a priori plausible, then the argument against a working memory explanation is invalid. Further, our own work suggest that working memory may account for at least a portion of the comprehension pattern attributed to patients with left frontal convexity lesions (see this post).

Broca's area is not specifically implicated in agrammatic comprehension. It is well known (well, maybe not well known, but well established) that damage restricted to Broca's area does not cause Broca's aphasia (Mohr, 1976; Mohr, et al. 1978). Agrammatic comprehension -- i.e., the pattern of comprehension deficits that Grodzinsky and Santi are trying to explain -- is associated with Broca's aphasia, not with Broca's area lesions. From this we can infer that lesions to Broca's area alone do not cause syntactic movement deficits. Add to this the observation that conduction aphasics, i.e., patients with posterior lesions, also tend to exhibit agrammatic comprehension, and you have a one-two punch: Lesions to restricted to Broca's area don't cause agrammatic comprehension and lesions to other brain regions (and sparing Broca's area) can cause agrammatic comprehension. From this we conclude that Broca's area plays no special role in syntactic movement (Hickok, 2000).

One of the complications here is that there are probably a number ways to cause deficits on the comprehension of sentences with long-distance dependencies. These structures tend to be harder to comprehend in control subjects. As such, one would expect that any disruption of processing efficiency, e.g., working memory or attentional deficits, could cause "impairments" in the comprehension of these types of sentences. Of course, a specific disruption of the syntactic operation computing long-distance dependencies could also disrupt comprehension on these sentences, but other possible sources of the deficit are rarely assessed, let alone ruled out when testing such patients.

We still don't understand the role of Broca's area in sentence comprehension, working memory, action processing, semantic processing, or any of its other possible functions.


Y GRODZINSKY, A SANTI (2008). The battle for Broca’s region Trends in Cognitive Sciences, 12 (12), 474-480 DOI: 10.1016/j.tics.2008.09.001

Roel M. Willems, Peter Hagoort (2009). Broca's region: battles are not won by ignoring half of the facts Trends in Cognitive Sciences, 13 (3), 101-101 DOI: 10.1016/j.tics.2008.12.001

Hickok, G. (2000). The left frontal convolution plays no special role in syntactic processing. Behavioral and Brain Sciences, 23, 35-36.

Mohr, J. P. (1976). Broca's area and Broca's aphasia. In H. Whitaker & H. A. Whitaker (Eds.), Studies in neurolinguistics, vol. 1 (pp. 201-235). New York: Academic Press.

Mohr, J. P., Pessin, M. S., Finkelstein, S., Funkenstein, H. H., Duncan, G. W., & Davis, K. R. (1978). Broca's aphasia: Pathological and clinical. Neurology, 28, 311-324.


Anonymous said...

I find this debate a little depressing, mainly because it reminds me of the futility of trying to chase the ghosts of cognitive psychology into one corner of the cortex.

This is one of my favourite review papers of the last few years. Hein and Knight propose that the function of the posterior STS varies according to task-dependent interactions with other areas of the brain. They only review data for the STS in the paper, but I would argue that the same principle should apply to any part of the cortex. Here's the abstract;

"The superior temporal sulcus (STS) is the chameleon of the human brain. Several research areas claim the STS as the host brain region for their particular behavior of interest. Some see it as one of the core structures for theory of mind. For others, it is the main region for audiovisual integration. It plays an important role in biological motion perception, but is also claimed to be essential for speech processing and processing of faces. We review the foci of activations in the STS from multiple functional magnetic resonance imaging studies, focusing on theory of mind, audiovisual integration, motion processing, speech processing, and face processing. The results indicate a differentiation of the STS region in an anterior portion, mainly involved in speech processing, and a posterior portion recruited by cognitive demands of all these different research areas. The latter finding argues against a strict functional subdivision of the STS. In line with anatomical evidence from tracer studies, we propose that the function of the STS varies depending on the nature of network coactivations with different regions in the frontal cortex and medial-temporal lobe. This view is more in keeping with the notion that the same brain region can support different cognitive operations depending on task-dependent network connections, emphasizing the role of network connectivity analysis in neuroimaging."

Hein G, Knight RT. Superior temporal sulcus--It's my area: or is it? J Cogn Neurosci. 2008 Dec;20(12):2125-36.


Greg Hickok said...

I agree that trying to pinpoint a function in one brain area may not be the right approach, but at the same time, I think Hein and Knight went to far in the opposite direction. I believe that a single brain region can be involved in multiple functions depending on the task, but there is a limit to this multi-functionality. Such a brain area must be doing *something* that makes it relevant to the tasks in question, and not to, say, olfaction. Regarding STS specifically, I think that if we looked at a finer grain of detail, we would find that even within the posterior STS there is functional differentiation to some degree.

Anonymous said...

I agree completely, but I'm not so sure that when we are talking about a single region we will be able to define that *something* in terms of the standard language of cognitive psychology. It still makes sense to me to use terms like 'phonological processing' when we refer to the action of a network of interacting regions, but not when we want to describe the functional competencies of a single patch of tissue.

Greg Hickok said...

Exactly right, Tom. I think it is highly unlikely that a box in a cognitive model will correspond with a postage stamp size chunk of brain. For example, the notion of "phonological processing" that David and I have been promoting is task-dependent and involves a different network of regions depending on these task demands, such as whether a mapping to meaning is required vs. a mapping onto articulatory systems.

David Poeppel said...

Remember V1 ....

Just to add my $0.02, when we interpret the putative 'function' of a single area, it will be helpful to remember the complex structure -- and, hence function -- of small patches of visual cortex. It's pretty unlikely, I think, that we can/should/will attribute single functions to tissue chunks -- that's just, pardon me for saying it again and again, the wrong granularity. There are likely numerous circuit-types in any chunk of cortical tissue, so what I take out job to be is to deconstruct the tasks into elementary computational subroutines.

And, to come to a boring conclusion, therefore both Tom and Greg are right: there are basic functions that do what they do (Greg), but they must probably be interpreted in context of their joint function across a network.

And what are the chances that our methods will succeed in identifying functions at the smallest scale? Ugh ... small, at best.