{% include_relative header.md %}
Representations are widely believed to be among the essential ingredients for explaining cognition. In cognitive psychology, behavior and decision making are modelled in terms of computations performed over representations. In neuroscience, responses of single cells, neural populations, and structural features are claimed to function as representations for the agent in question. In AI, what machines learn is often characterized in terms of the kinds of representations those machines acquire and how they manipulate them.
However, although the term 'representation' makes frequent appearances, there is little agreement about how it should be understood in either neuroscience or AI.
The aim of this special issue is to obtain views of leading philosophers, neuroscientists, computer scientists, and individuals from other cognitive science disciplines about how we should answer (one or both) of the following questions:
- How should we understand talk of representation in the brain?
- How should we understand talk of representation in AI systems?
Given the recent use of AI models to describe brain activity, and the increasing adoption of large, hard-to-interpret models in AI, we expect there to be some productive interchange between these two questions. Authors may, however, wish to focus exclusively on one domain.
The aim of the special issue is specifically to showcase positive views about how to understand representation in these domains. Authors are encouraged to stake out ideas that are bold and constructive. Articles should also be clear and well argued (with the understanding that a positive position on this topic might not be immune to every possible objection). We are deliberately soliciting less guarded contributions on the topic than are often found (hence the 'bold'). We think that there is a value for readers to see a variety of opinions on how to tackle the big questions defended by a diverse range of experts in the field, and the special issue as a whole will serve to map out the 'logical geography' of the field for readers. Please note that we welcome papers that are critical of either whole or part of the representational approach. We hope that the special issue will form a key reference point for future work on how to understand representation in these domains.
Examples of possible questions a contributed paper might address include:
- Why is it useful or necessary to ascribe representations to the brain and/or AI systems?
- How should we distinguish between different types of representation?
- What evidence should we draw on to ascribe representations to the brain or AI systems?
- What is the relationship between representation and learning?
- How should we make sense of the notion of a vehicle of representation for these domains?
- Does representation talk in these domains pick out an objectively real feature in the target system or is it only a useful construct for external theorists?
- How might we naturalize representational content?
- How should we understand ascription of probabilistic representations to subpersonal systems?
- What connections are there between the methodology for finding representations in the brain and for finding representations in AI?
- How might the history of representation and our use of representation in other contexts color its current use in neuroscience or AI?