Synthese (12):1-34 (2018)

Cameron Buckner
University of Houston
In artificial intelligence, recent research has demonstrated the remarkable potential of Deep Convolutional Neural Networks (DCNNs), which seem to exceed state-of-the-art performance in new domains weekly, especially on the sorts of very difficult perceptual discrimination tasks that skeptics thought would remain beyond the reach of artificial intelligence. However, it has proven difficult to explain why DCNNs perform so well. In philosophy of mind, empiricists have long suggested that complex cognition is based on information derived from sensory experience, often appealing to a faculty of abstraction. Rationalists have frequently complained, however, that empiricists never adequately explained how this faculty of abstraction actually works. In this paper, I tie these two questions together, to the mutual benefit of both disciplines. I argue that the architectural features that distinguish DCNNs from earlier neural networks allow them to implement a form of hierarchical processing that I call “transformational abstraction”. Transformational abstraction iteratively converts sensory-based representations of category exemplars into new formats that are increasingly tolerant to “nuisance variation” in input. Reflecting upon the way that DCNNs leverage a combination of linear and non-linear processing to efficiently accomplish this feat allows us to understand how the brain is capable of bi-directional travel between exemplars and abstractions, addressing longstanding problems in empiricist philosophy of mind. I end by considering the prospects for future research on DCNNs, arguing that rather than simply implementing 80s connectionism with more brute-force computation, transformational abstraction counts as a qualitatively distinct form of processing ripe with philosophical and psychological significance, because it is significantly better suited to depict the generic mechanism responsible for this important kind of psychological processing in the brain.
Keywords abstraction  connectionism  convolution  deep learning  empiricism  mechanism  nuisance variation
Categories (categorize this paper)
Reprint years 2018
DOI 10.1007/s11229-018-01949-1
Edit this record
Mark as duplicate
Export citation
Find it on Scholar
Request removal from index
Revision history

Download options

References found in this work BETA

Thinking About Mechanisms.Peter Machamer, Lindley Darden & Carl F. Craver - 2000 - Philosophy of Science 67 (1):1-25.
A Treatise of Human Nature (1739-40).David Hume - 1978 - Oxford University Press.
A Treatise of Human Nature.David Hume & A. D. Lindsay - 1958 - Philosophical Quarterly 8 (33):379-380.

View all 50 references / Add more references

Citations of this work BETA

Understanding From Machine Learning Models.Emily Sullivan - 2022 - British Journal for the Philosophy of Science 73 (1):109-133.
Making AI Intelligible: Philosophical Foundations.Herman Cappelen & Joshua Dever - 2021 - New York, USA: Oxford University Press.

View all 25 citations / Add more citations

Similar books and articles

Some Neural Networks Compute, Others Don't.Gualtiero Piccinini - 2008 - Neural Networks 21 (2-3):311-321.
Abstraction Relations Need Not Be Reflexive.Jonathan Payne - 2013 - Thought: A Journal of Philosophy 2 (2):137-147.
Out of Their Minds: Legal Theory in Neural Networks. [REVIEW]Dan Hunter - 1999 - Artificial Intelligence and Law 7 (2-3):129-151.
The Nuisance Principle in Infinite Settings.Sean C. Ebels-Duggan - 2015 - Thought: A Journal of Philosophy 4 (4):263-268.
The Troubled History of Abstraction.Ignacio Angelelli - 2005 - History of Philosophy & Logical Analysis 8.
Structural-Abstraction Principles.Graham Leach-Krouse - 2015 - Philosophia Mathematica:nkv033.
Marx et les abstractions.André Tosel - 2002 - Archives de Philosophie 2 (2):311-334.


Added to PP index

Total views
715 ( #11,054 of 2,533,576 )

Recent downloads (6 months)
53 ( #15,445 of 2,533,576 )

How can I increase my downloads?


My notes