Talk:Tactile object perception
Review
I found this a beautifully written and highly informative article. I have no specific suggestions for changing it, but I do have a couple of pointers to potentially relevant work.
(1) The discussion of more detailed modeling of exploratory procedures could encompass other work (e.g., Knut Drewing's analysis of force control); however, it will be covered in the entry on haptic exploration by Klatzky and Reed, so maybe a pointer to that is in order.
(2) The discussion of neuronal conjunction raises the important point of whether there is convergence on a single "grandmother" cell or whether complex features are coded by neural populations. The authors might look at the paper of Abraira and Ginty (Neuron. 2013 Aug 21;79(4):618-39. doi: 10.1016/j.neuron.2013.07.051.) on the role of the dorsal horn in modulating the initial inputs from mechanoreceptors.
These comments aside, the paper is good to go as it is. It is a very fine contribution.
Response to Prof. Klatzky
A reference to the 'Haptic Exploration' entry has been added. The Abraira and Ginty paper was left out as it appears to us somewhat far removed from the topic at hand, in a paper which is already quite dense and extensive in scope. The authors wish to thank Prof. Klatzky for her informative remarks and kind words.
Response to Prof. Sathian
This is a reply to Prof. Krish Sathian's insightful remarks to this paper, which were edited directly into the article on 02.21.15
The following is the reply of the authors. For the actual changes made, one can compare the the current version (03.11.15) to the one made edited by Prof. Sathian (02.21.15), using the following link: [1]
1. Comment: In relation to covariation of shape and texture, individual differences may be important: in both vision and touch, object imagers tend to integrate surface characteristics such as texture into their shape representations, whereas spatial imagers do not (Lacey, Lin & Sathian, Exp Brain Res. 213: 267-273, 2011). This might explain some of the inconsistencies between studies. Reply: We accept this informative comment and have incorporated it into the paper, by adding the following sentence in the relevant place: “Other explanations are, of course, possible. For example, it was recently shown that different cognitive styles (namely, the tendency to employ different types of imagery) may affect the co-processing of texture and shape, suggesting a meaningful role of individual differences in these tasks (Lacey, Lin & Sathian, 2011).”
2. Comment: With regard to the cortical coding of texture, work more recent than the cited papers has shown a role for parietal opercular as well as visual cortex in tactile texture processing (Stilla & Sathian, Human Brain Mapping 29: 1123-1138, 2008; Sathian et al., NeuroImage 57:462-475, 2011; Eck et al., 75: 123-135, 2013). Further, there is evidence supporting dual pathways in touch as in vision, with segregation of the object properties processed by each pathway (Sathian et al., NeuroImage 57:462-475, 2011). Reply: We accept the comment regarding the opercular cortex’s involvement in haptic texture processing, and have added the following sentence in a relevant place: “Indeed, recent studies have found further support for the involvement of the parietal opercular cortex in haptic texture processing (Stilla & Sathian, 2008; Sathian et al., 2011)”. We also accept the comment regarding the possibility of multi-sensory cortices in processing haptic information. However, it appears discussion of these subtleties may hamper the focus of the current article, which is already quite detailed. As such, we feel a short referral is the best approach, and have added the following sentence: “One intriguing suggestion has been that some of these features may be processed by non-somatosensory cortices, specifically the visual cortex (Sathian & Lacey, 2015)”. As for the comment regarding the possibility of dual pathways, mention of this idea has been eliminated from earlier versions of this paper for 2 reasons: 1. It does not necessarily add information to the current topic (feature information should still be clearly evident in ‘what-pathway’ areas, and this is the matter at hand). 2. More importantly, it conflicts with evidence, which appears to us stronger, favoring a representation of action and object combined, outlined in section 4 of this paper and dealing with active perception.
3. Comment: replaced “processed by a specific individual primitive” with “corresponding to a specific individual primitive”; replaced “processed by tactile primitive” with “processed as tactile primitives”. Reply: We attempted to use the term primitive to refer to a complete unit of processing, as opposed to a simple input feature of the system. This use of the term may appear slightly unconventional, as the term usually simply refers to the features on which the system constructs its operations. We chose to employ the term in such a way here based on the consideration that features may sometimes be inseparable from processes, a notion at the heart of our section on active sensing. We acknowledge that our use of the term ‘primitive’ might have been slightly too liberal, given its status as a “reserved term”. We have thus replaced ‘primitive’ with either ‘processing primitive unit’ or ‘primitive unit’ when referring to processing primitives (as opposed to primitive features). The word “primitive” by itself may refer to both feature and processing units, according to the context.
4. All other minor remarks were accepted.
Guy Nelinger, Eldad Assa, Ehud Ahissar
