Spacecog scribbles May-Jun 2007

03.05.2007 – on crossmodal space. Neurons respond to spatial information from more than one modality. See Spence & Driver (eds). Some are “multimodal”, others are mostly unimodal yet respond to changes in other sensory modalities. Spatial attention also displays multimodal attributes, whether the attention is endogenous (“voluntary”) or exogenous (triggered by “outside” stimuli). So again we have here the issue of our phenomenology (separate sensory channels), and the more complex reality (multimodal processing both at neuron and neural network level).

Then there is also the fundamental issue of how it is that we build the various partial spaces, and how the different centricities are to be characterised. We think of our spatial world as unified, but the various centricities and the “scaffolding” of information storage (using the world as a storage space for spatial information and bringing various spatial-centrisms to make use of parts of it when required) make this a complex theme. The motor impact and input of/to learning can be taken even further though also with machines and virtual machines – cf. Phillipona, O’ Regan et al on a robot which learns by doing that the world (as well as it’s world) is n-dimensional [dd - ms available from internet, no other ref info to hand]. Cf also Quartz & Sejnowski’s “Neural Constructivism” (1997, 2002). [<- from “remarks on haikonen”].

Macaluso & Driver, “Functional Imaging of Crossmodal Spatial Representations”, in Spence & Driver (eds), Crossmodal Space and Crossmodal Attention (p266):

“We.. find crossmodal effects of spatial attention on areas traditionally considered to be unimodal (i.e. on occipital areas that respond to visual stimuli and that show more activity when vision is task-relevant). The fact that we now find this even for a case of endogenous spatial attention, where attention is directe voluntarily, impies a role for top-down modulatory signals (presumably sent by backprojections) in relating spatial information about one modality (e.g. which hand to attend for a tactile task) to areas that are primarily concerned with another modality (vision here). “

Also in Spence & Driver there are articles considering the different egospatial-centricities (body-centred, head-centred, retina-centred, gaze-centred, hand-centred). To what extent do these interact with “other-centred” spatial paradigms such as “in this box-centred”, other-person-centred, which we also seem to use. Other-centring and other centring decisions seem to be prime function of language understood in the cognitive linguistic paradigm, perhaps this is why Grush is such a fan of Langacker. See esp Langacker ch2 "Inside and Outside in Cora” in Concept, Image, and Symbol.

07.05.2007 – writers like to give “pointers for future research”… a kind of infallibility-I’m-the-best-game where each author seeks to give the strategic directions for others to follow gladly. This can be seen in a variety of philosophical and soft-science (linguists, human evolution) texts. Whereas I think future research should follow the principles as correctly established in my book, perhaps this should become a standard text for all...

14.05.2007 – On Alon.. ch 3 negative autoregulation is under selection for rapid response-times to protein production to a steady state level (homeostasis), works better (faster, more robust, more control then unregulated production to a set value) cf standard deviation against random network occurrence. Used for Positive autoregulation is slower but more “fate-determining” therefore useful in developmental trajectories.

There is good evidence for convergent evolution in selection of FFLs e.g. Z and Z´homologous, but X,X´and Y,Y´non-homologous. Thus evolution is strongly constraining transcription factor networks to use functional FFLs which maximise usefulness of two environmental inputs Sx, Sy. Two inputs are under selection as they increase the dynamic control options for protein production.
15.05.2007 – see Alon p93 for summary of sensory transcription network motifs, inc also SIMs (single-input-modules), multi-output FFLs, and DORs (densely organised regions), including their main physiological functions.

Joëlle Proust’s article mentioning Evans and Strawson on space.

Space and cognitive neuroscience; spaces egocentric and other. Systems biology and spatial knowledge. E.g. chemotaxis. Movement and cell biology – flagella; muscles and neurons. Phenomenal space and connectionism. Animal ego and allospaces.

04.06.2007 – Sole & Goodwin p136 – Freeman’s idea that the brain goes from a state of chaotic attractors to a (lower-dimensional, categorical-perception cf Gärdenfors) state of limit cycling when perceiving e.g. odors.