What is a forward model? And what might forward models be useful for?
How, if at all, should discoveries about motor processes and representations inform
attempts to solve The Problem of Action?
- Wolpert (1997) but please skip the last section on motor learning (which we will not discuss).
Before you read this, try out some very small-scale actions while paying attention
to the way your arm, hand and fingers moves:
place a finger on the spacebar of your keyboard and then move it to (i) a near key; and (ii) a far key;
do the same with a small obstacle blocking your direct path near the target key; and
use your finger to trace out a figure of eight (∞)
with a small loop and a larger loop.
using a mug or cup with a handle, grasp it around the top, lift it up, rotate it (i) through 15 degrees or (ii) through 45 degrees then place it down.
If you are short of time and need something shorter, much of the above is contained in:
If you want more, consider any more recent work by Wolpert, Kawato and collaborators.
Or, if you can get hold of it:
This seminar’s question also requires reading set for previous seminars:
Where to Find the Reading?
In some cases the reference section of the lecture notes already
includes a link to help you find the reading.
If there is no link in the lecture notes, start by searching for the title
(and, if that fails, by title and authors) on google scholar. If this fails, the library has resources. If those fail, please check first with others on the course. If you still have problems, you may email your seminar tutor.
Ask a Question
Your question will normally be answered in the question
session of the next lecture.
More information about asking questions.
: A model used to predict ‘the sensory consequences of a motor command’ (Wolpert et al., 2003, p. 595). In action performance,
forward models are used to compensate for feedback delays,
to distinguish self-produced movements and their sensory consequences, and to
select between different courses of action.
: The ‘process by which the ... outputs of the [motor] system are specified given an extrinsic task goal.’
(Wolpert, 1997, p. 210)
The Problem of Action
: What distinguishes your actions from things that merely happen to you?
(According to Frankfurt (1978, p. 157), ‘The problem of action
is to explicate the contrast between what an agent does and what merely happens to him.’)
Butterfill, S. A., & Sinigaglia, C. (2014). Intention and motor representation in purposive action. Philosophy and Phenomenological Research
(1), 119–145. https://doi.org/10.1111/j.1933-1592.2012.00604.x
Davidson, D. (1971). Agency. In R. Binkley, R. Bronaugh, & A. Marras (Eds.), Agent, action, and reason,
(pp. 3–25). Toronto: University of Toronto Press. https://doi.org/10.1093/0199246270.001.0001
Frankfurt, H. G. (1978). The problem of action. American Philosophical Quarterly
Rosenbaum, D. A. (2010). Human motor control
(2nd ed.). San Diego, CA, US: Academic Press.
Wolpert, D. M. (1997). Computational approaches to motor control. Trends in Cognitive Sciences
(6), 209–216. Retrieved from https://doi.org/10.1016/S1364-6613(97)01070-X
Wolpert, D. M., Doya, K., & Kawato, M. (2003). A unifying computational framework for motor control and social interaction. Philosophical Transactions: Biological Sciences
(1431), 593–602. Retrieved from http://www.jstor.org/stable/3558137
Wolpert, D. M., Ghahramani, Z., & Jordan, M. (1995). An internal model for sensorimotor integration. Science
(5232), 1880–1882. https://doi.org/10.1126/science.7569931