End of 2007 comment

More a formality than a contentity this blog post, with it being the end of year but without me feeling in much of a review mood. Think one of my 2008 tasks will be to read through these posts, a self-indulgent activity for which Andy Thompson and me coined the word "nurnation" and it's verb form "to nurnate" (i.e. to read through something you wrote yourself). Thanks to all you millions of readers and thanks to blogspot for helping me see more clearly now via postings.

Commentous moment, momentous comment

Thanks O. So in 2007 I do at least get a comment (the previous one was a test one I did, so it doesn't count). Hurray for me, and you. Now I need to find your email details, since you are not on Facebook.

Or should I comment on your comment, that's probably more sensible... done.

Read an excellent book on the evolution of cognition, "How Homo became Sapiens" by Peter Gärdenfors - who is a genius. Such difficult material, presented so well.

Bechtel's book on cell biology is an excellent approach to the philosophy of science, emphasising the importance of, well, mechanisms and diagrams in scientific explanation.

Customer review of Haikonen's book on Robot Brains can be found here:
[Click here for book details and review] - I sent a more detailed and technical review to the author, he seemed happy with it so everyone's happy then, which is nice.

No rest for the nerd oh well

Been a bit quiet on the spacecog front lately - i blame facebook, moving to london, and British Telecom. Now then, back on message...

Bechtel and Abrahamsen's "Connectionism and the Mind" manages to consider connectionism both historically/up-to-the-present (place in cognitive science etc), and philosophically. Also the widest ranging discussion of connectionism i've read. Excellent.

Now working through Haikonen's "Robot Brains". Not that he is a robot or anything. Maybe a biobot, like the rest of us. Anyway I like this book - a synthetic approach to machine cognition; lots of clear diagrams and explanations of his associative mechanisms; considers many aspects of cognition; and has some sense of the philosophical issues that surround his chosen subject. More detail, and more interesting focus overall than in his 2003 book (<-which is maybe more suitable for Humanities faculty undergraduates reading around a new subject).

"Spatial Cognition II" (Freksa et al)

Very brief, somewhat impressionistic remarks on these papers posted to Google Docs, see link in sidebar.

Overall - some good papers to refer back to, some off message for me. 80 euros a bit steep for what I got, but never mind.

Mike Rinck's paper: "Spatial Situation Models" 2005

in Shah & Miyake: "The Cambridge Handbook of Visuospatial Thinking". Excellent article. Emphasises the spatially cognitive aspect of language-comprehension, and the exploratory, fragmentary, active, goal-directed nature of imagining spatial situations from descriptions.

General remarks on cognitive psychology... Based on the various articles, the psychological level of explanation is directed at around the level of "behaviour attractors" in thinking. This is a useful level, even though I get the impression that these researchers find the phenomenology of thought as confusing as I do. Some researchers seem addicted to the "mental images as pictures" model, but not all, e.g. Tversky, and Rinck, who proposes associative mechanisms in spatial situation model construction.

Spatial cognition and the psychological perspectives

Taking a bit of time to adjust to this psychology style of thinking about space (Shah & Miyake's book). Had to skip a few chapters (bit lazy there, motivation a bit elsewhere, must be london calling and the footie starting up again), and even the good ones need a bit of wood-chewing (Barbara Tversky article excellent, Newcombe & Learmonth good).

Interesting this continuing robots-n-philosophy dialogue with Haikonen. Mirrors, he's been doing a smoke and mirrors game for the AAAI symposium. This time I think an email will do, only a cupla points to raise there and anyway need to collect my writing strength a bit for the ol' "to-be-started".

Thompson on egocentric and allocentric space

see p.391 - emphasises the need to recognise others as having egocentric spaces in order to conceive of oneself as inhabiting an intersubjective space.

Review of Haikonen's article completed and sent, well done me. let's see what he thinks about it.

Remarks on a few articles

The journal "Phenomenology and the Cognitive Sciences" has some downloadable articles from Vol 6 (2007)... cue downloadings varieuse... Siewert, Noë, Thompson, all concerned to discuss Dennett's heterophenomenological method; and Dennett's reply to various authors from that issue.

As far as i can see, others criticise Dennett for his de-emphasis of first-person reporting as a possible methodology, while Dennett, typically, seeks to correct what he sees as the various misinterpretations of his method, with his own brand of gentle needling of opponents combined with imaginative "what about this?" scenarios.

Particularly enjoyed Dennett's version on the "I am great"-storyline ([c] Dylan Drummond, Andy Thompson, and men throughout the language phase of our natural history), where in conversation Ned Block says "The letters seemed blurry" as a first-person report and Dennett replies: "So, did they seem to you to be difficult to read because they were blurry, or did they seem blurry because they were difficult to read?" - goddim!

Also Dennett makes an amusing remark about how our cultural and intersubjective backgrounds may affect a supposedly controlled experiment - delays in word completion, where the words presented are "cun-" and "shi-".

Now then, Haikonen's article... plenty of relevant issues there to be considered... next week's essay project.

more musings on thompson

glad to see he's made a mistake in Ch. 5, falling back on outdated "five kingdoms" biology classification. but then he goes on to mention Niche Construction and Odling-Smee... that's encouraging, we converged on that independently (i reckon, etc).

---

Need to get those Merleau-Ponty notes onto cyberpaper sometime. probably about 40 pages though.. hmm.

Initial musings on Thompson's "Mind in Life"

Only on page 70 just now, but clearly this is one of the best philosophy books ever written.. it's like reading "the book i would have liked to have thought i could write".. still, chin up, he's leaving room for the magnum opus on subjective space. Dynamic-systems theory, enactive approach, phenomenology, neuroscience, co-evolution of organism and environment... pressing all the right buzzerwords and references for me.

Also good to see Thompson acknowledging Stephen Priest's exegesis of Merleau-Ponty (see Ch3 Note 1.). Stephen Priest: a man who has done a lot of thinking. Hope he's well.

--- [jumps to sth else via various associative connectionisms...]
Sth i thought i had written down somewhere but hadn't, it seems: in Sterelny's book, the idea of co-evolution and in particular of niche construction is important, perhaps i should get that Laland/Odling-Smee book on the subject

Brief comments on articles in "Naturalizing Phenomenology"

Like the articles in Petitot et al. Dagfinn Follesdal "Gödel & Husserl", and Giuseppe Longo: "mathematical intuitionism and phenomenology". Mathematics has to be realised in a non-Platonic manner. Also links into Quine, to Gärdenfors conceptual approach, and even Wittgenstein's philosophy of mathematics.

Also good was the introduction to Husserl and naturalisation by the editors; Petitot's article (complex, theory-laden, but interesting); Tim van Gelder's article, Varela's article (both on time). the Thompson Noë Pessoa article quite good, though familiar and overall somewhat programmatic in the end.

Excellent commentary on Husserl and modern cognitive psychology and subjectivity by Maria Villela-Petit. Also good for Infallibility down the pub with The Generalissimo would be the last article (Jean-Pierre Dupuy), a look at the philosophical history of early and later AI and it's relevance to the debate about naturalising or even doing phenomenology.

Quotes and comments on Morgan's "The Space Between our Ears"

P65-6 – “Motion is a specific sensation from a particular class of neural mechanism, just as colour is a sensation. We cannot divide sensation of motion into separate pictures like the frames of a motion picture, because those sensations were never there to start with. Motion is computed directly from the image, without sensory intermediates."

Idea here is that since there are brain cells in visual cortex which respond to motion (in preferred directions), so we can forget the empiricist ideal of constructing motion from sense-data.

P98-99 – “Generative models are also beginning to explain how we move our muscles. Moving our limbs has much in common with interpreting our retinal images. In both cases the number of possibilities is so vast that we have to simplify them with models. There are roughly 600 muscles in the human body. Even if we assume for simplicity that each muscle can be contracted or relaxed, there are still 2 raised to the power 600 possibilities – more than the number of atoms in the Universe. It does not seem plausible that every possible movement we might wish to make is represented in the brain by a memory of exactly the right signals to be sent to each individual muscle. The alternative is that we in some way represent the desired end state of the movement – an internal model – and have neural networks that we tune by experience to translate these high level models into action. If action involves the comparison between a model and the present state of the body, the distinction between perception and action begins to fade away."

Note also that the teleological models are also tuned through sensorimotor experience e.g. in karate.

P128 – “..we say that the mirror has reversed left and right, but we are mistaken: the reversal has been done by our act of mental rotation”

- similar solution to Nerlich and Martin gardner – the enantiomorph is rotated through 4-dimensional space.

---

Other quotes:

Preface, ix: "The prize for the worst poem about science has yet to be awarded... The zoological lines of John Hookham Frere in 1799 have... received a mention:

'The feather'd race with pinions skims the air
Not so the mackerel and still less the bear' "

P19 "Wyld's 1815 'Chart of Civilisations' colours countries from 'I' (very uncivilised) to 'V' (England). Canada gets an unkind II for containing 'cannibals and Frenchmen': it is not clear whether it was the cannibals or the French who raised it above Australia (I)."

P150 "The next section examines the idea that only some areas of the of the brain are involved in our conscious experience. Readers should skip this section if they think that the relationship between philosophy and science is like that between pigeons and statues"

Musings from last Night of the Newts

sleepy sleep, come to me.... ach, no yet... what about then, a sort of running commentary on the text, by the "author"... can give it a bit of structure by outlining (in a sidebar at the start of each chapter?) the plan of action, boxes'n'links style like say... could be good to point out "author-sensed-weaknesses" as we go through. Like the idea of having boxes and side-notes or footnotes in the text. These should appear near the main text, not at the end of the chapter, that's always a bit annoydart flickin forward and back aw ra time... mmm, sleepy sleep, descending like a giant grey fog... a giant grey frog... reddit.. there must be already a book written which is basically a preface "i intend to talk about... thanks to... " with no actual main text... why do i like that idea? is it even my own? how do we know what we have thought up ourselves and what is just a forgotten reference? and what is just putting a few more bricks in the edifices created by others... ahh, sleepy sleep, röh röh, zzzzzzz.

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.

Some notes on spacecog from 2006

11.07.2006 – to answer the question of what thinking is, we need to synthesise neuroscience, the evolution of language, and both the origin and subsequent most relevant) evolution of life.

Rose (“lifelines”): Genes are not privileged in evolution – also genomes, RNA, ribosomes, evo-devo processes, the co-evolving environment are part of the evolving organism.

Co-evolution is a fundamental part of evolution – thus the environment is not passive, and the organism is not passive in its interaction with the abiotic and biotic environment.

14.07.2006 – wittgenstein’s points about qualia (see Dennett – consns expld)… non-explanation of “private” language… sense-data a non-starter for theory-building, these aren’t describable… a point still missed in most philosophy papers to journals… “we define the model as a set of predicates SP such that statements P(L) map onto experiential states E of the world” etc etc. (my old dissertation contained some awareness of this).

18.07.2006 – memory as a stored pattern of synaptic weightings in neural networks. Similarity of declarative memories and perceptions, though perceptions always considered stronger e.g. Hume.

When considering neural models of perceptual representation and memory, difficulty that we are considering circuits with 10x neurons and 10y synapses. -> “grainyness” of resulting models when we think of our own experiences

26.07.2006 – tensor network theory and connectionism as theories showing how sets of neurons can map phase spaces using tensors etc (in Churchland – neurophilosophy).

15.08.2006 – Stiny on what constitutes a boundary – differences between points and higher dimension entities as boundaries. Lack of existence of any predefinable atomic vocabulary for shapes -> impossibility of counting (using rules to define a linear series). Fractals are unfinished but in their definition they are determinate (the rules are fixed, only the results lead to indeterminacy of boundaries).

25.08.2006 – not surprising that most of our language involves descriptions of movements or positions in a space, since motor understanding is the “basis” of “higher” cognition. See lakoff on metaphors, grush on egocentric vs allocentric space (in churchland, “brain-wise”).

06.09.2006 - Alva Noe and visual perception as process (temporal), like somatic perception. Avoid reification into “pattern/state”.

10.09.2006 – when the empiricists’ logical atomism failed, the mistaken way to deal with it was to think that philosophy “ended” with Wittgenstein, the philosopher who provided both the most mesmeric expression of logical atomism (in the Tractatus) and some of the most scathing, though somewhat gnomic, criticisms of that atomism (in Philosophical Investigations). The best way to deal with the end of empirical atomism is to know enough about the science of perception to be able to begin philosophising about perception again.

Although many neuroscientific experiments and actors are labouring under some old Cartesian and atomistic illusions, not all of them are, at least not all of the time. Bennett and Hacker’s book is depressing because of their deliberate narrow-mindedness on this point, combined with the repetitive “Wittgenstein was right – philosophy analyses how ordinary language is used”, as if thinking about perception or doing science was basically a subsiduary side-issue for philosophy.

25.09.2006 – evolution of language and communication. Cellular and molecular comms and the naturalistic turn to the concept of “information”. Neural comms and computation. Bioinformatics and mathematics – naturalism in mathematical logic. And complexity. Genetic switches. UV, biosonar, bioluminescence and biofluorescence etc animal sensation/perception and philosophy. Bug eyes (compound) and other eyes. A phenomenology of bee vision?

Animal comms and philosophy – see Lieberman also. Mimesis and meaning, and M-P. Lieberman p201 meaning and use. Macdonald 1994 reference in Lieberman. Connection between cognition, motor ctrl and language.

29.09.2006 – self/non-self e.g. at cellular, molecular, organism levels. Self-organisation - molecular /astrobiological/evo devo/neural. Fractal (repeating) aspect of levels in biology. Hox genes and dnA – how we’re all (nearly) the same – the biologic approach to philosophical issues. emergence of consciousness issue – biological aspect to mind/body problem. Problem of the interspecific boundary between types of consns or cons/non-consns. Perception - Need to broaden the enactive view out here, as Clark says in Is the vis..?

30.09.2006 – chance and necessity and evolutionary process. Contingency of existence – talkorigins, gould, monod, Sartre etc. biological processes and aesthetics (fractals, Fibonacci series, ramachandran on art).

01.10.2006 – on drosophila eyes: http://xray.optics.rochester.edu/workgroups/cml/opt307/spr04/greg/
Konrad Lorenz institute: http://www.kli.ac.at/

02.10.2006 – issue of whether experience is continuous or gappy. Epistemological problem with both views is that we can’t experience “gaps”, and neither can we experience continuity. Wittgensteinian point – we don’t have anything to contrast the continuity with. The “Stream of consns” model is an inadequate metaphor. If consns is like a stream at, it’s more like a stream which has turned into an ocean seafloor current. It envelops its experiencer completely.

But Dennett, Ballard et al have good points against the “standard continuous” model, since through science we have shown that we are not continuously perceiving. The issue is to then explain why it doesn’t “seem gappy” if it isn’t “really continuous. Llinas tries to fill the gaps with the oscillatory “closed system” model.

Phosphenes – find out more about. The nearest thing to pure qualia around. 03.10.2006 Phosphenes are an entoptic phenomenon – see wikipedia http://en.wikipedia.org/wiki/Entoptic_phenomenon. Journalofvision.org, vision research (Elsevier) many articles available online

04.11.2006 - Entoptic phenomena are in some sense the “sensation” experiences, but as with other “sense-data” they remain fairly ineffable and are described always by perceivers who have the non-entoptic capacities and categorisations.

The problem for the animal of differentiating, and the problem of integrating, the sensory channels. How many sensory channels are there. Is there – can there be – more than one current perceptual channel, or if the sensory channels are perceptually integrated, what does this integration consist in?

09.11.2006 – [sth about p267 re Kirk Ludwig in the book "contemporary debates in cognitive science".]