Ted Krueger’s Paper Proposal

Classification of Enactive Interfaces

[Traditions 3,7,8]
The shift from a computational to an enactive understanding of perception is a fundamental one that has opened new possibilities for interfaces, prosthetics, tele-robotics, sensory substitution, and data ‘visualization’. This perspective enables a diversity of applications while simultaneously uniting them in a common framework that allows for a transfer of technologies, techniques and insights across the range. The purpose of the proposed paper will be to systematically structure the relationships between the various fields of application in order to understand those that are actively researched and those less so; to understand both accomplishments and opportunities.
This interest developed when I was asked to consider the application of sensory substitution techniques to problems related to diabetic neuropathy. Presently I am engaged in a research and development project that will attempt to substitute manufactured pressure transducers for biological ones and to interface their output to portions of the body unaffected by the neuropathy. This is not sensory substitution as developed by Bach-y-Rita and others in that the tactile is not standing-in for the visual, for example. However, it is what I have termed sensory translocation, and the underlying principles are similar.
From a sensory standpoint this neuropathic deafferentation renders the biological limb similar to a prosthetic one and suggests that there may be a direct transfer of technique from the diabetic to a prosthetic context, that this tactile translocation technique should be effective on a prosthetic limb as well. In addition, I am also engaged in a project to give a tactile sense to a mobile robotic arm being developed for a paraplegic. In this case the effector is remote and independent from the body, but again the strategy is closely related. Here are classes of augmentation all of which might share a common approach, indeed the strategy in these research projects is to move developments and insights rapidly across each of these applications.
The enactive approach to perception and cognition is closely related to and builds on the informational closure of the organism. It is this that suggests additional applications, in simulation and ‘visualization’. If perception relies on structured variation between afferent and efferent signals it should be possible to simulate the operation of a robotic hand for example and not rely on the physical hardware that is in any case remote from the body (and perhaps out of sight). This may be useful in training and technology development. However, once the transition to a data stream has been accomplished, it will be possible to substitute any structured data source. The reference to, in the present example, pressure transduction could be substituted for similarly organized dimension of a dataset regardless of the supposed correlation to the physical – what I have referred to as data perceptualization.
While much work remains to complete the proposed classification, I believe that giving a structured overview of the potential for cross-fertilization between these diverse research projects will yield insights that advance the research area.

Cybernetic traditions:

  • 8) Neurobiology; consciousness studies
  • 3) Experimental epistemology; constructivism; philosophy of science

4 thoughts on “Ted Krueger’s Paper Proposal

  1. Thomas Fischer

    Great to see that ASC members are looking at prosthetics! There was a number of papers on neural prosthetic work at the recent conference on Norbert Wiener, who, as you probably know, also worked on prostheses. The recent work presented was mainly conducted by a group at the University of Washington. It included ideas about digital security in neural prostheses and about giving patients some level of learning/feedback control over the performance of neural prosthesis via mobile apps. Fascinating!

    I find it hard to imagine how you will classify work/projects/systems/principles in this field and look forward to seeing your presentation.

  2. Ranulph Glanville

    I an deeply intrigued, specially since I know your background and have seen it growing and developing. It’s wonderful to see such a range of previous interest build into this project.

    What I see as interesting and also difficult about classification systems is that they apply and gradually enforce an order that it becomes hard to transcend. Consider the way we have defined our senses as 5, and the difficulty we have even increasing that to 6 (kinetic sense), let alone to considering other ways of grouping and categorising. The result is something very conservative and, while enlightening, also arguably restrictive. I will be very curious about how you approach this. Your abstract is cagy. I hope your classifications aren’t!

    1. Ted Krueger

      I agree with your concern about classifications. I don’t see my classification as in any way absolute, but only a way, among potentially many ways, to structure a field. What I hope to show is as much the connection between the categories as the differences. I’m interested in the transfer of insights or techniques between them.

  3. Peter Cariani

    This is a very interesting line of exploration.

    The diabetic neuropathy system sounds like it is remapping the insensitive cutaneous territories into still functional ones. One of the advantages in these in-modality remappings (as opposed to the cross-modal sensory substitution) is that the neural coding scheme is preserved and that existing sensory regions in the brain can more easily adapt (via a deformation) in a manner that is going on all the time in virtually all parts of the brain. (IMHO, one of the problems with many cross-modal sensory substitution attempts was that they did not get the neural coding scheme right — they assumed that spatial patterns of receptor activation were adequate and left out the fine temporal structure of the stimulus and receptor responses).

    In terms of neurophenomenology, I think that the boundaries of what we perceive as parts of ourselves follow divisions between internal circuliar causations and external contingencies. We percieve as parts of ourselves everything that our brain reliiably controls, everything else that appears contingent to us iwe regard as part of the external world. When we wield prosthetic devices (e.g. a poking stick), the functional boundaries of our selves move out to the distal end of the stick (where the point of contingency lies). So whenever there are strong action-percept correlations, there is the possibllty of internalizing and automatizing those processes — when things become completely predictable, they become automatisms that no longer are experienced consciously as sensations.

    I think that you are right in thinking that the correlation structure is the key. Correlation structure is what differentiates the modalities. As the song goes, “it’s not the meat, it’s the motion”.


Leave a Reply

Your email address will not be published.