Of snakes, the pulvinar, and fear

A new paper in PNAS suggests that “Pulvinar neurons reveal neurobiological evidence of past selection for rapid detection of snakes” (from the title). I’m happy that more research is being done on the functions of the pulvinar, a structure that is fascinating. There are many interesting findings in the paper, and it’s certainly worth reading.


The problem, as usual, is not with the results but with their interpretation.Establishing selectivity to visual stimuli is challenging at best (cf. all the disputes re. faces in ventral visual cortex). Some puzzling (and to me telling) aspects of the data that the authors barely discuss are:

  • Good responses were observed to high spatial frequency stimuli (!), not just low pass images. In fact, the effect of low vs. high pass had a small effect size (given a p value that was only < .1)
  • Latency to snake pictures was fast (around 55 ms on average) but how much faster than other stimuli it was not clear (but maybe I missed this).
  • The authors suggest that they recorded from the medial pulvinar (the “associational” sector). Talking to colleagues who are familiar with the intricacies of pulvinar anatomy in several species, the  figure shown by the authors does not make this point convincingly. The authors really need to demonstrate that this is not visual pulvinar (that is, from what is shown it is not clear that they were in the medial pulvinar as described in the literature).

These are issues that can be resolved with further research. My main concern is the evolutionary conclusion of the paper.  As phrased by the authors: “Our data provide unique neuronal evidence supporting the hypothesis that snakes provided a novel selective pressure that contributed to the evolution of the primate order by way of visual modification”. This is unfortunate; I’m not a comparative neuroscientist, but without studying multiple extant species, a claim like this is clearly over-reaching.

Reference: Van Le, Q., Isbell, L. A., Matsumoto, J., Nguyen, M., Hori, E., Maior, R. S., … & Nishijo, H. (2013). Pulvinar neurons reveal neurobiological evidence of past selection for rapid detection of snakes. Proceedings of the National Academy of Sciences, 110(47), 19000-19005.

Low road vs. multiple roads

The processing of affective information has many attributes that make it special, such as speed, and relative independence from attention and awareness. A key question, therefore, both from basic and applied perspectives is how this happens. An extraordinarily popular account is that a so-called low road from the retina via the superior colliculus and pulvinar conveys information to the amygdala. The general idea is that, because the pathway is entirely subcortical, processing would then be automatic.

This proposal has captured the attention of the research community and has fostered several lines of investigation — what is the role of attention, of awareness, how fast are certain effects, what type of visual information is conveyed (low vs. high spatial frequency), etc. Although these questions are interesting, the subcortical pathway idea is, in my view, largely based on an idea, rather than solidly grounded on empirical data.

So for a while now, Ralph Adolphs and I have been discussing what are serious problems with the notion of automatic subcortical processing of affective information. We have now written up some of these ideas in this Opinion piece in Nature Reviews Neuroscience. We also propose a new scheme, called the multiple-waves model that is intended to be an alternative to the “standard view”. It looks like part (B) of this figure, in contrast to the more traditional view shown in (A).

The proposal also incorporates the fact that the pulvinar is a highly integrative thalamic region, with extensive interconnectivity with much of cortex, as shown below.

The pulvinar works in a way that integrates cortical-subcortical processing.