The important part of Lupyan’s theory is that the effect of language on thought takes place online — it does not create long-lasting changes in cognition or perception (which is why it can be disrupted by aphasia). This is in contradiction to previous theories that have been used to support the idea of linguistic relativity, which is what makes the label-feedback hypothesis so interesting.

This is the third post in a three-part series on Gary Lupyan’s label-feedback hypothesis. Before getting into the interesting psychology of the label-feedback hypothesis, I would like to suggest that you read the previous two posts in the series, Is Linguistic Information Part of Every Cognitive Process? and Language Interference and Cognition.

Now that you are up to speed on the theory, I would like to point out why this is such a big deal. If the label-feedback hypothesis holds, this means that linguistic processing, unless it is disrupted by aphasia or an experimental condition, is active in just about every cognitive process. And if that is true, the line between linguistic and non-linguistic cognition has been significantly blurred, if not altogether annihilated. Exactly what that means is open to interpretation, and it is awfully complicated to think about.

The mechanism that Lupyan proposes for this linguistic involvement is an interactive cognitive processing model, in which different mental processes interact to create the phenomena that we consider under the umbrella term “cognition.” And if he is right, then thinking and learning could be much more complex than we initially imagined. While I am not an expert in cognitive models, I would think that trying to pin down any specific effects or phenomena would be significantly more difficult if we assume that any number of processes could be affecting each other simultaneously and automatically.

Of course, this could also affect how we do cognitive research. If researchers can come up with a theoretically viable account of which processes are interacting, we could, by process of elimination, support or refute the idea through interference-based experiments. This can be seen in the Lupyan’s experiments that I alluded to previously — he had a theoretical hypothesis that language affected categorization, he used interference to eliminate the linguistic effect, and he saw that categorization was performed differently.

Finally, and possibly most importantly, the label-feedback hypothesis has something very interesting to say about how concepts are stored in our brains. Lupyan holds that there is no core concept of a category; for example, there is nothing in our mind that holds the prototypical essence of a dog. Instead, whenever we are making a classification decision about dogs, we combine previous knowledge — dogs we have seen before — with current task demands, which could include telling the difference between dogs and cats, or dogs and foxes, or dogs and other dogs.

In this way, the label-feedback hypothesis actually proposes a very new kind of cognition: One that is modulated by multiple online processes, including language, and is affected by knowledge gained in the past and action undertaken in the present. I am still trying to fully understand and conceptualize this, as it is a significant departure from classical theories as well as many modern ones. Lupyan and others will certainly be putting a lot of time and energy into working out whether or not this is a viable model for human cognition, so watch for more news in the near future.

As I continue to study Lupyan’s ideas and those of others that think similarly, I will try to keep you updated on what I find out. And if you have any suggested readings or potentially useful information, please share!

References

Lupyan G (2012). Linguistically modulated perception and cognition: the label-feedback hypothesis. Frontiers in psychology, 3 PMID: 22408629

Image via OrganAlle / Shutterstock.