Engineering conceptual change: The Enactive Torch

1
Engineering conceptual change The Enactive Torch
Adam Spiers Bristol Robotics Laboratory Department
of Mechanical Engineering University of Bristol,
Bristol, UK
Ron Chrisley and Tom Froese Centre for Research
in Cognitive Science Department of
Informatics University of Sussex, Brighton, UK

• Workshop on Philosophy and Engineering 2008
• Royal Academy of Engineering, London
• Tuesday, 11 November 2008

2
Overview

• This meeting
• Not just Philosophy of Engineering
• But Philosophy and Engineering includes ways in
  which Engineering may assist philosophy

3
Engineering forconceptual change

• Some problems are not empirical, but conceptual
  philosophy
• Solving conceptual problems sometimes requires
  not just manipulation of concepts already
  possessed, but acquisition of new concepts
• Acquiring the right concepts may require more
  than just thinking acting/experiencing
• Design and construction of devices may assist in
  providing the kinds of experiences required
• Thus, engineering may assist philosophy in
  developing new tools (concepts) for solving
  conceptual problems
• Example The enactive torch

4
Example the enactive torch

• Simple distal-to-tactile sensory substitution
  device
• Translates the distance measures of one
  ultrasonic sensor to a single tactile (rotary or
  vibratory) output to the hand

5
Example the enactive torch

• Engineered for the purpose of solving conceptual
  problems in the phenomenology of perception

6
Three ways to engineer for conceptual change

• Design loop Design and build artefacts that do
  X so that the experience of designing itself
  produces new concepts of X (et al)
• Use loop 1 Design and build artefacts the use of
  which produce new experiences of Y, that in turn
  prompt new concepts of Y
• Use loop 2 Design and build artefacts the use of
  which produce new experiences of Y, that in turn
  prompt new concepts of experience itself (Z)

7
Philosophy assolving conceptual problems

• Philosophy provides methods for conceptual
  analysis and development
• (Focus in this lecture is on the method of
  analytic philosophy, or at least what it is
  conventionally believed to be)

8
Analysis is propositional

• Problem solving within the analytic method is
  (taken to be) exclusively propositional
• Assumes a static stock C of basic concepts
• Emphasis on creation of new propositions out of C
• If new concepts are proposed, these are logical
  combinations of concepts in C

9
The limits ofpropositional analysis

• Solving some conceptual problems requires
  concepts not in C, nor equal to some logical
  combination of concepts in C
• If so, then solution of these problems requires
  methods not currently taken to be part of
  analytic philosophy

10
The limits ofpropositional analysis

• E.g., the mind/body problem can't be solved with
  only our current concepts of mental and physical
• "We may hope and ought to try as part of a
  scientific theory of mind to form a third
  conception that does directly entail both the
  mental and the physical, and through which their
  actual necessary connection with one another can
  therefore become transparent to us. Such a
  conception will have to be created we wont just
  find it lying around." (Nagel 1998)

11
Extending the analytic method

• This is not to say that the required new methods
  are not philosophical
• Since these methods will have the function of
  providing the right concepts for resolving
  philosophical, conceptual problems, it is right
  to see them as philosophical
• Rather, the current view of the method of
  analytical philosophy, either as it is, or as it
  could be, is incomplete

12
Beyond concept empiricismInteractive Empiricism

• Concept empiricism
• The acquisition of (some) concepts requires
  having (certain kinds of) experience
• Interactive empiricism
• Concept empiricism, plus
• The acknowledgement that the required experiences
  are typically interactive
• The experiences are not just sets of input, but
  a dynamic coupling between action and perception.
  (cf Held and Hein)

13
Concept acquisition asnon-propositional activity

• Concepts are skills, and and at least some skills
  cannot be acquired propositionally, in the sense
  above
• (E.g., Cant learn to ride a bicycle solely by
  reading about it.)

14
A role for engineering in philosophy

• Proposal Engineer devices that permit new kinds
  of interactive experience, which in turn enable
  acquisition of concepts essential to solving a
  philosophical problem
• Not just technology, but iteration of design/use
  of the technology engineering

15
Three ways to engineer for conceptual change

• Design loop Design and build artefacts that do
  X so that the experience of designing itself
  produces new concepts of X (et al)
• Use loop 1 Design and build artefacts the use of
  which produce new experiences of Y, that in turn
  prompt new concepts of Y
• Use loop 2 Design and build artefacts the use of
  which produce new experiences of Y, that in turn
  prompt new concepts of experience itself (Z)

16
Designing for (new) experiencethe enactive torch

• The design phase of the current prototype was
  (is) a cycle closely coupling first-person
  experience and engineering
• E.g., initial output was a rotational disc which
  measured distance by turning under one's thumb
• Resulting experience of using the device was a
  detached, reflective attitude directed toward the
  position of the disc
• Since this was not what was sought, re-designed
  the device to provide more immediate involvement
  with the world continuous vibrational output

17
The enactive torch andconcepts of perception

• Conceptual problems in the philosophy of
  perception
• E.g. "Is perception independent of action?"
• Traditionally Yes
• Enactive theories of perception No
• Latter can be hard to grasp, understand, or
  motivate
• Engineering to the rescue

18
The enactive torch andconcepts of perception

• Tom Froese
• "When demonstrating the enactive torch at
  conferences, I initially tried to explain the
  basics of the enactive approach to perception
  first, before handing over the device.
• This was because it was originally conceived as a
  device to inform that particular debate in the
  cognitive sciences, and most people had never
  even heard about the main tenets of that
  approach."

19
The enactive torch andconcepts of perception

• Tom Froese
• "However, it soon became clear that this
  theoretical introduction was not only
  unnecessary, it was even confusing to most
  people.
• How do you convey a cognitive science research
  program in a few sentences to someone who has no
  idea of what cognitive science even is?
• It was only after using the device that many
  people had an "aha!" moment - now they could
  grasp the idea that embodied action is important
  for perception."

20
The enactive torch andconcepts of perception

• Tom Froese
• "So eventually, I hardly explained the device at
  all, and just gave it to people to try out for
  themselves.
• In this way they obtained a much better
  understanding of the essence of the enactive
  approach, in their own lived experience, without
  me having to explain any theoretical background
  at all."

21
Enactive torch design constraints

• Portable (non-intrusive, low energy requirements
  can take to conferences!)
• Intuitive (minimal training)
• Doesn't require vision (would interfere with
  desired phenomenological analysis)
• Throw away colour information focus on distance
  (ditto)
• Resolution of output
• high enough to enable sensory substitution
• low enough to enable discovery of the role action
  plays in determining richness of experience
• Low cost

22
Engineering conceptual changeToward an
empirical study

• Proposal Empirically measure the extent to which
  experience with a sensory substitution device can
  change ones concepts of perception
• Method Ask subjects to indicate their degree of
  assent to statements about perception and action
  before and after use of enactive torch
• Controls use of normal torch (and reading
  philosophy texts about perception?)
• Similar to experimental philosophy, but emphasis
  on conceptual change, and engineering

23
Beyond philosophy of perception

• In some cases, may even be best if the
  philosopher is the designing engineer
• E.g., Tom Froese and the enactive torch
• Impetus of conceptual change not limited to
  experiences that engineered devices enable
• Rather, experience of designing the artefact may
  itself prompt the conceptual change required
• Problem much harder to measure
• But The proof of the pudding

24
Thank you

• References
• Chrisley, R. (in press). "Interactive empiricism
  the philosopher in the machine", in McCarthy, N.
  (ed.), Philosophy of Engineering Proceedings of
  a Series of Seminars held at the Royal Academy of
  Engineering. London Royal Academy of
  Engineering.
• Chrisley, R. (2008). "Philosophical foundations
  of artificial consciousness", Artificial
  Intelligence In Medicine 44119-137.
• Froese, T. Spiers, A. (2007). Toward a
  Phenomenological Pragmatics of Enactive
  Perception, in Proc. of the 4th Int. Conf. on
  Enactive Interfaces, Grenoble, France
  Association ACROE, pp. 105-108.
• More information on the enactive torch is
  available at http//enactivetorch.wordpress.com
• Comments welcome
• ronc_at_sussex.ac.uk