Back    All discussions

Linking action to semantic memory
I'd appreciate some feedback on a speculation concerning the relation of action theory and the physiology of memory in the brain. There has long been an intuition in historiography that a consciousness of the past is a keystone of liberty and in biology that an organism''s novel action depends on memory or hysteresis. In the last two decades, there has been rapid progress in understanding the brain, and I'm trying to reconcile it with action theory.

I get the impression that today memory refers to an emergent system effect---that is, a result of complex interactions between different brain areas. Memory apparently falls into two broad classes: semantic memory and an implicit memory. The latter appears to be is a set of rules (habitus) for stimulus-response relations. Further, I get the impression that both classes of memory are constructed as emergent effects of sensation and learning and are essentially static in that changes in them are extrinsic in origin (the possibility of long-term epigenetic biological inheritance seems marginal to the question I'm raising).

If this is a fair representation of current views, then the problem I need to address is that, if action is always innovative in the sense that it produces a novel state of affairs, one that is counterfactual in relation to semantic memory, then there has to be a constraint upon semantic memory that makes the circumstances represented in it acquire possibility (an alternative state of affairs is possible) and potency (change is possible), or their combination as a probability distribution of possible outcomes of action in relation to the state of affairs represented in semantic memory.

It seems to me this sense of possiblity and potency can't be embedded in the executive function, for some knowledge of a probability distribution is not consciously learned. For example, pre-conscious learning about causal relations from the haptic sense. This suggests to me that possibility and potency are ontologically real rather than just epistemological tools. That is, I'm not inclined to embrace the curious Enlightenment doctine that the mind has a supernatural creative power that is ontologically independent of external determinations, as an Olympian capacity to generate hypothetical counterfactuals, among which it freely chooses, perhaps employing epistemological criteria such as tests or saving the data.    

I'd appreciate knowing if it is possible that knowledge of potency and possibility might be acquired by the class of implicit memory that establishes rules to constrain how intentional action links with semantic memory. That is, action must draw upon semantic memory to represent the context for action and it looks to the executive function to focus that action, but it also might look to implicit memory for the rules that add a probability distribution to the circumstance represented in semantic memory. Is this possible, reasonable, or probable?

I have had trouble finding literature related to how the ontological implications of action might be handled in terms of brain functionality. 

Haines Brown 

Linking action to semantic memory
Reply to Haines Brown
Dear Haines, I doubt you will find anything on 'ontological implications of action' in accounts of brain function, even in the work of people like Patrick Haggard, Dan Wolpert or Tony Marcel, who deal with the relation between control of action and sense of action (in the scientific sense of motor activity). It is difficult to see quite what your ontological implications are but I worry that they relate to concepts that cannot be cashed out in the language of natural science (like 'liberty'), if indeed they can be cashed out at all. Your 'possibilities' appear to be clearly epistemic rather than dynamic/ontological. The neural computations making use of these judged possibilities will of course ground them in real brain dynamics, in the form of short and long term synaptic reinforcement as part of memory but I am unclear in what way this raises new ontological issues. Robots will do the same sort of thing and we understand enough about their ontology to build them. I am also unclear why possibilities cannot be embedded in 'executive function'. Inasmuch as I understand what is meant by executive function the work of Wolpert and Haggard would suggest that possibilities are indeed embedded there. Why would that be a problem?  


Linking action to semantic memory
Jo, I apologize for my question being rather vague and open-ended.

One problem with it was my presumption that possibility and potency are ontologically real, not merely epistemological artifacts. That a scientific realism in support of such a position has become broadly held in the philosophy of science and may be part of an emergent consensus is illustrated by Boyd, Gasper and Trout, The Philosophy of Science, which is considered in these climes to be a standard old text. Of course, those in the analytic and phenomenological traditions probably object, but at least today a scientific realism is neither unconventional nor inherently suspect (what is going on in Oxford may be incomprehensible in Cambridge).

Another problem is what is meant by information transfer, say, from properties of the world, to sensory apparatus such as the retina, conveyed though a neural network to the visual cortex, and finally becoming the consciousness of a scene. It seems conventional to speak of this as a causal chain in which the relation of links (closed entities) is that of analog, model, function, mark transfer, energy transfer, correspondence, reflection, etcetera. More fashionably, one might say that the observable properties of each link emerge from those of the prior on which it supervenes. However, this only describes a constant conjuncture of observables and is not explanatory.

While one might think of the retina of the eye as being struck by photons of certain frequencies and power, but to be technical the photon is only a probability distribution that is actualized when it becomes framed, when it enters a relation with the retina or with observational instrumentation. The signal from the retina to a nerve does not consist of photons, but is an electro-chemical signal that weakly varies in relation to the probability distribution of the photons. Indeed, it seems the retina constructs an electro-chemical  signal that is arises from own capacities, and it is its own particular actualization of the photon probability distribution that constrains the signal it generates.

Doesn't this scenario rather challenge the conventional distinction of ontology and epistemology? Indeed, while I'm an ignorant outsider, I have the impression that the model of the brain in terms of complex causal relations of independent modules is currently rather in question.

The issue I raised was about our grasp of the real potency and possibilities of our world. In conventional phenominological terms, the world we experience can consist only of static observables, and so "process" can only be a constructon of mind. Action, however, must "see" the world as having real possibilities and potencies and know them as real. Surely these properties are real in fact, for otherwise change in the world would never occur.

I am reminded of St. Augustine's comment about time. He very well knew what it was, but was incapable of describing it to someone else. I am also reminded of discussions of the haptic sense that conveys a pre-conscious (relatively independent of memory) knowledge of causality. It seems that the problem is really the inability of the mind to represent process in conscious thought because the categories of thought and language depend on static memory (all traces of the past are static).

You observe a flowing stream and see that it is flowing, and it is not just a hypothetical construction of conscious thought. While what constrains the possibilities of the sensory apparatus must be the actual structure of the world, this actual structure is in principle always a becoming rather than a being (however, in the back of my mind is something about sensory apparatus sampling input information rather than respond continuously. But out of ignorance I put this aside).

So my question was that because memory is apparently a static structure and, as you point out, the world as process has to be reconstructed in conscious thought, I was trying to locate the location of that representation outside memory, and I raise the question of whether that representation is only a hypothetical reconstruction or if instead it is directly experienced and known. That is, the mental apparatus involved in activity seems to know the actual world is process rather than having merely to infer a past process from a sequence of memories.

Sorry for all the speculation here, but I'm trying to reconcile neurology with my understanding of what is emerging in the philosophy of science.

Haines Brown

Linking action to semantic memory
Reply to Haines Brown

Dear Haines,

I think I can see where you are coming from and I would agree that there are questions of this sort that need attention. I do not think you will have much luck with the literature though! At present neuroscientists have nothing to offer in response to the question ‘what neural event corresponds to the sense of making a decision’ or even 'what makes memory semantic'. I have ideas of my own but they involve some serious lateral thinking.

I am sure that ontological possibility is ontologically real, but I doubt this has much to do with the epistemic possibilities used in thinking. Thus, if the Deep Blue computer is asked to ‘decide’ what move is most likely to win from a chess position it will do so, based on epistemically possible/probable moves by the opponent. However, those possibilities will not be modeled by ontological probabilities. The computer 'action' will be pre-determined by the programme: probability 1. If the ‘opponent’ is in fact Deep Blue itself running alternate moves then the whole game is pre-determined despite complex computations on epistemic ‘possibilities’. I suspect our brains are not so different.

I entirely agree that the popular mode of description of biology (I would forget emergence, which to me is a codeword for b*llsh*t) begs deep questions and that the conventional concepts of ontology and epistemology are naïve and even back to front. However, I do not think that this alters the fact that what are conventionally called ‘epistemic’ and ‘ontic’ probabilities share very little apart from an assigned value from 0 to 1.

I am not sure that ‘process’ is a construct of mind any more than ‘observation’. To my view the world we experience is likely to consist of concrete ‘actual goings on’ or processes that determine our experiences. We cannot experience these goings on as such but we run goings on in our heads that model them and allow us to understand them enough to make useful predictions.

However, I am intrigued by your idea that ‘action must see the world as having real possibilities’. I think I would strongly agree with that. The difficulty I see is that the brain has evolved with very complex and counterintuitive transduction relations between ‘outside events’ and the inside events that model them. To use an example I used in my book, I think the sense of good and bad relate to basic physical events. In a sense ethics is grounded in physics. The trouble is that the sense of good and bad arise in the model event in the brain and George W Bush and Tony Blair have shown that what seems good or bad inside has such an indirect relationship to what might be good or bad outside that ‘internal ethics’ are no guide to world ethics. (Even the electrons in Deep Blue might sense good and bad, but again, with no relevant relation to winning or losing.) Similarly the spurious idea of change as making something happen that ‘was not going to happen’ has little to do with change in the pervasive sense of physical dynamics, and even our sense of movement is not in fact motion etc. etc.

There are some serious issues here, I agree, but I am skeptical that the philosophers’ idea of ‘action’ is a useful starting point. Moreover, I do not think received neurobiological dogma can currently handle these questions because it shies off the very things you would like it to address.


Linking action to semantic memory
Jo, thank you for the interesting reply.

As for the word "emergence" being b*llsh*t, I agree in a sense. Because I found it problematic, I once looked a bit into the history of the word in the 19th-20th century, and found that it was originally used rather loosely, perhaps the equivalent of novel observable properties. However at some point (such as Pepper, 1926), a basic distinction was introduced between mechanistic accidental contingencies and non-mechanistic unpredictable outcomes. This had the effect of restricting the word henceforth to epistemological unpredictability. This hides issue of whether the reason for unpredictability was ignorance (situations too complex to predict such as perhaps the n-body problem) or the presence in an initial state of unobservables (hidden variables such as the "functional" properties favored in the social science or unobservable extrinsic properties often brought up in the natural sciences). So I use the word in its old generic sense of objective novelties present in any process, which in principle all things are. So while the word looses significant content, I nevertheless find it useful to emphasize that one always addresses change a priori rather than presume static entities and then make change post-facto..

> I am not sure that ‘process’ is a construct of mind any more than ‘observation’. To my view
> the world we experience is likely to consist of concrete ‘actual goings on’ or processes that
> determine our experiences. We cannot experience these goings on as such but we run goings
> on in our heads that model them and allow us to understand them enough to make useful
> predictions.
This in a way was my question. If memory is a structure, as such it seems incapable of encoding process, for a (closed) structure by itself and without injecting some free energy is intrinsically static. On the other hand, the trend in neuroscience seems to be away from a model consisting of causally connected modules, in which case "memory" might not simply reduce to structure. For example, if "memory" is defined as having essential extrinsic properties, then it would seem to be essentially a process that is not static (I'm grossly simplifying my actual position on this matter). The issue here would seem to be the implications of the word "essential". It might mean that there can be no persistent memory without its being simultaneously connected with other regions of the brain. Do you have any thoughts about these things? There's work suggesting that accessing memory necessarily changes it, and in terms of my own scheme, that might be because memory could be reprsented not as a structure, but as a probability distribution that is framed by other parts of mind and the actualization of its possibilities thereby constrained by those other parts. 

I agree, that the world we experience is a world of processes, but the issue is whether the "experience" is itself a process. The rushing stream I'm looking at is a process, and the retina I use to I see it is receiving, not data bits of discrete properties, but, say for the sake of discussion, vectors.

Now, there seem to be two possibilities: a) the retina cannot respond to the directional component of the vector, but just the values of its magnitude, and so it reduces the input to, say, a rapid sequence of magnitudes. If so, my clear conscious impression that the stream is a process must be an unconscious mental re-construction of process, not an experience of process. b) the electro-chemical mark (in this case, the wrong word for it) produced by the retina somehow can represent vectors, not only their magnitude but also their direction.

Is this possible? A vector is just a mathematical construct, and so what does it mean in physical terms? I have some electronics background and so instinctively imagine a case such as this: a wire connected to an oscillocope lies in a moving magnetic field, and to judge by the display on the scope, the wire does in fact transmit a process. Although the sweep of the scope can only sample the magnitures of a process, I believe one can safely infer that the input signal is continuous, the physical equivalent of a mathematical vector. That signal is not a sequence of discrete values being sampled (unless _possibly_ one gets down to the quantum level), for the values are continually changing and are never specific magnitudes, but smeared out in time. Based on this analogy I see nothing wrong with the notion that the neurons transmit a process rather than discrete data bits.

However, most of what goes on in my visual cortex and especially my conscious mind as I observe the moving stream is drawn from memory, even though I think of myself as simply observing the moving stream in the actual present. There seem to be a range of possibilities: a) if memory is not defined as a closed entity, but as a process, then there would seem to be no problem about direct experience of processes; b) if on the other hand memory is a static affair, then enough vector info might reach the visual cortex independently of memory to enable me to see the stream directly as a process (you seem dubious about this: "We cannot experience these goings on as such"); c) in cognitive ethology, animals have such hard-wired capabilities as seeing and recognizing things as entities and at least in mammals to distinguish self and other. Perhaps a knowledge of process is a hard-wired capability to interpret changing discrete input data as process; d) on the other hand, if the info coming into consciousness consists of discrete magnitudes, since birth we have subconsciously learned to understand these discrete inputs as processes.    

My question was aimed only to find out the extent to which these possibilities are possible or probable. You point out that there's not much literature concerning such matters and that it perhaps raises unresolvable issues.  And yet, the choice made here seems to have far reaching implications for methodology. For example, if one does a brain scan to measure some value for a region of the brain, one implicitly sees the region as a closed entity having a self-contained causal power or disposition (there are problems with both, for ultimately neither are observable instrinsic properties: a causal power is the relation of an actual state and a possble and more probable state; a disposition is a function of boundary conditions).  

My aim was not to cram neurology into my favorite conceptual hobby horse, but rather to see if that scheme is at least not-incompatible with current brain science. Although I've had to dig into the philosophy of science and my current work is on social theory, these are foundational for work in my own field, which is historiography. My object is to construct a conceptual scheme and methodology in historiography that is philosophically sound, even if not entirely conventional, that is physical and compatible with the various sciences, including neuroscience and social science.

You are not enamoured by starting out with the philosopher's idea of "action'. I'm not sure of your implication here. If in context it only refers to how one goes about doing neural science, being an outsider that is not my concern. However, in the social sciences it does make an enormous difference. For example, the conventional notion of action is that intentionality is a causal factor rather than a constraint on the world's possibilitities and potentials. This presumption generates the insoluble dead end of semantic truth value, such as in a correspondence theory. My approach is to insist that the primary object is to change the world, not just to understand it. Intuitively I suspect that neuro science might also be quite different if the brain were understood as seizing upon the possibilities and potencies of the world rather than being an an autonomous causal agent with causal powers and dispositions. But given my ignorance and distance from the field, such an intuition has absolutely no merit.


Linking action to semantic memory
Reply to Haines Brown
Dear Haines, In posing the question 'are these ideas compatible with empirical neuroscience' I could give the simple answer 'no'. It has to be your option (a) for the retina and your impression of the stream as process must be a mental construction. (Hence a film of a flowing river gives the same impression of flow although there is none.) And I think that is the right answer to the question posed in the terms you lay out. However, there are premises underlying your terms that I would question. 
If I then ask why do I question them and who else would question them I find myself back in my slightly peculiar position of a 'fly up above a dividing wall' between two squash (fives, wall-game) courts in which the 'two cultures' of physical science and philosophy/social science play out. (There are other ways to cut the cake but this will do.) The issues you raise are to me fascinating and crucial to an understanding of man in his world. By and large neuroscientists spend their time solving detailed practical problems but refuse to admit even the existence of these issues, with the result that they cannot construct any coherent model of how the mind works. On the other hand the philosophers tend to grasp the issues but equally fail to make any progress because they misunderstand how the scientist uses his tools and build ideas that incorporate a 'folk physics' that physicists themselves discard. As a metaphor, the neurologists examine a sculpture wrapped in bubble wrap from outside the bubble wrap with high powered microscopes. The philosophers tear off the bubble wrap but do not even bother to put on their spectacles so neither recognises what the sculpture is of. Having worked in biomedical science for 30 years and finally decided I needed to remove the bubble wrap what I seem to find inside is a bit peculiar. It is a sculpture upside down and inside out with a clear inscription G L Fecit 1714. That is to say that everything in the conventional scientific canon works fine, but only if you are prepared to frame it in the seriously counterintuitive way that Leibniz did. So, my way out of your quandary is a very personal way in some senses.
Perhaps the key point I would make in response to your suggestions is that I do not think there is any distinction between a dynamic process and a static state. A static state is just a very slow process, i.e. a dynamic process in which certain values are more or less zero. This is not a trivial point. Those like Worrall and Ladyman who say there is only structure imply dynamic structure. 'Static states' are almost certainly just heuristic tools we use to link dynamics with experience in a way that cancels out the role of neural dynamics. If I perceive a house there is a complex stream of dynamics from sun to brick to retina to brain. 'House' allows me to discuss the dynamics before the retina and specifically the rather slow dynamics of some bricks. Dynamics can be just as efficacious as process no matter how slow (i.e. static) they are. The true complement to the dynamic process is the experiential state.
Secondly, signals or records cannot carry with them 'process qualities' from their last port of call. Since all world dynamics form a chain back to the big bang this would mean that every signal would carry qualities of all the previous processes this history. It generates an absurdity. 
Thirdly, our 'impression' of process is of course not a process but the result of a process. I am sure that we are born hard wired to perceive flowing streams as flowing etc. I suspect that the main difference between us and animals is that we are more loosely wired so that over our long period of infancy our impressions can be moulded to current social and environmental needs. We have words, but these do not bring new impressions. They just label them. Thus I suspect that animals are hard wired knowing what death is and that they will die. We learn at least the detailed probabilities of that because we are softer wired, but we cannot learn what death is. I think we are born with a fear of permanent oblivion and a belief that such a situation is at least a possibility.
We are also born with the idea that we can 'change things', which is useful but wrong!


Linking action to semantic memory
Jo, thanks for your remarks.

Just a brief comment about the "two cultures". Traditionally and as commonly understood, this distinction was between a model of the physical sciences (the positivist textbook laboratory model taught to students), and the social sciences, which deal with processes that are creative or "emergent" in sense of outcomes for some reason being only partially predictable. I've never been happy with this dichotomy, in part because it reduces the physical sciences to Newtonian mechanics on one hand, and it allows to slip in a mysterious Urgos on the other.

There are, of course, a range of physical sciences that deal with marginally predictable phenomena, such as geology, meteorology and cosmology. Put simply, it seems to me that Newtonian or mechanistic processes are merely an effect of closure, and no system can be absolutely closed. Indeed, it seems the fashion in the philosophy of science to see closure as a only a hypothetical limiting case. If so, then the physical and the social sciences are points on a continuum characterized by the degree of relative closure, not categorical opposites. To use a word that you with some justification criticized, everything is to some extent an emergent process. So we seem to agree that a static process is just a very slow dynamic process.

In my own field (historiography), it is conventional to presume stability or order and then to inquire as to reason for a change. But in light of the above, it seems one might better presume change and seek to explain the reasons for instances of its relative absence, for order or stability. I myself look to a structural theory of persistence (relation of constituents is probable), but some object to it. However, my sense is that an a priori presumption of change characterizes biology (although I'm not at all sure). For example, if "life" were to be defined as a combination of reproduction and metabolism, then it seems that life is a process to begin with, and one seeks the explain the order represented by an organ, organism or species. Traditionaly, random mutation is presumed and is not the object of explanation, but instead one looks to the constraints on the possiblities it offers, the selection that gives explains specific structures. 

I would also like to reflect, if you allow, on your remark about "signals or records cannot carry with them 'process qualities' from their last port of call", for this statement might reflect some misunderstanding of my position. I suspect it is because of the limitations of conceptualization and communication that we tend to reify Aristotle's categories of possibility, potency and necessity (actuality). However, I tried to suggest that, ontologically speaking, these are not separable things, but merely epistemological aspects of one thing, a deconstruction of process in thought. That is, I suggested not only that everything is a process, but I defined process as a probability distribution that combines actuality (local structure--the past), potency (which I technically characterize as extrinsic--the present) and possibility (which I technically characterize as exogenous because every process represents a constraint by its structure on a more universal process from which it originated--the future). Without dragging you into a swamp of theory, I define each aspect in terms of the others, and no one is sufficient unto itself. Therefore, a signal or record does not in fact reduce to its observables, to its actual structure, but is a probability distribution actualized as a new structure and thus a new probability distribution when the process of which the signal is an aspect enters into a relation with, is framed or localized by, another process, such as a detection apparatus.

An absurdity that I'm a scion of the Big Bang? Perhaps, but perhaps not. In what sense do I carry forth the Big Bang? Broadly, everything that exists represents a constraint upon constraints upon the possibilities that the Big Bang brought into being. Not only am I ultimately a constraint on its possibilities, but also on its potency (assuming the universe is winding down toward an ultimate heat death---an assumption now in contention). In thermodynamic terms, the probability gradient that enables my activity ultimately derives from the probability gradient first created by the Big Bang, although that gradient is necessarily smaller, as the Second Law insists.

And this brings us around to your friend Leibniz. Is it possible I'm trying to offer a physical explanation of his theodicy? More relevant than it might at first seem, in, say, the 4th to 7th centuries A.D., the aesthetics of the classical Mediteranian world yielded to a new asthetic. The former found meaning to be resident in surfaces, while latter found meaning to arise from relations. For example, in the 10th century, the word "pauvritas" did not mean a lack of material wealth, but the lack of a necessary social relation, and for a while friendship was sacramental. Although living in a period of transition, I suspect Leibniz still held to the traditional (feudal) scholastic world view. However, the dominant trend (there have always been oppositonial  movements), in the 17th-18th century was toward the position that meaning is intrinsic to entities, arise from its constitutents and the forces acting on them, such as a concern for defining human nature or the reduction of meaning to observables (and I would perversely argue, still in functional explanations). Sorry, I really intended to keep away from such idle speculations.


Linking action to semantic memory
Reply to Haines Brown
Dear Dr. Brown,

I felt immediatly attracted by your question. Please, forgive me if I'm not giving any comment to what is discussed above. Many good and true statements have been made but I think the whole question is much broader. You link two highly complex phenomena in your question; "action" and "the semantic memory". Action, on one hand, direclty deals with the mind-body problem and in concrete the question about the existence of such a thing like "free will" or "agency". In your discussion we can find this reflected in questions of determinism of the world, for example. The semantic memory depends on the definition of time. Will and time are deeply interrelated subjects.

Since I am working on this question only for 2 year now in preparation of my doctoral disertation, my contribution will not be scholarly. I can rather give you some inspirations by my own mistakes. The following ideas may help you to approach the problem from a different angle.

1) First, and this is rather personal. The more you enter historiography of Enlightment and related earlier or later work, you will see that it is difficult to lable any philosophy according to its time period, school or tradition. The question you make have been dealt much more controversaly within each of these schools.

2) Second, I have come to the point that, as you rightly state, these issues deal not so much with epistemology or ontology but with the conection between both. We can often find comments that relate or oppose these two fields. But I believe, it is a misconception. Rather than a dialectic problem it is a triangulation between epistemology, ontology and ethics. I fear, it is imposible to make any conclusions of the suggested relation between will and time, knowledge and meaning, subject and world, science and reality without entering the field of ethics.

3) As for the problem of semantic memory and action, I suggest you to read Alfred Schutz and his Phenomenology of the Social World which he builds on Mises and Bergson.

4) As for the problem of subjectivity and freedom of action, I recommend to enter into the field of psychology and psychoanalysis. I found much useful work by Slavoj Zizek and of cours Viktor E. Frankl.

5) The only one I know, who links these two fields is Albert Schweizer, especially in his book "Reverence for Life".

6) A short remark at the end, you rightly speak of semantic memory as a kind of mental representation (signs, etc.) of past occurences in the individual mind. There are two things, I did not find in your question, which may be for reasons of being short.
         A) Memory is not only concerned with the past, it is also concerned with the future. Then we call it vision. The simple belief that the past has any relevance now, depends on the idea that we can project into the future. You will find this much more detailed by Schutz.
        B) Memory, and this is perhaps the most challenging we learn from cognitive science, works different from individual to individual. It represents a past as it is remembered, already codified in the sign system or meaning-context of the individual which through communication is actualized.

7) I found some interesting inspiration by the contrastation of "counterfactual truth" with "paraconsistent logic". All your questions are addressed within this field of logic.

Feel free to contact me any time on these questions, I feel they are to broad for a deeper discussion here. If you like, you may have a look at my most recent working paper, where I struggle with a very similar question than you: WP Performative contradiction

You may find it useful to start with "Part II" where I make some notes on "logic" and the different understanding about it.

Good luck and best regards.
Tabea Hirzel

Linking action to semantic memory
Reply to Tabea Hirzel
Ms. Hirzel,

Your comments were interesting and very much appreciated, and I took a quick glance at your paper, but I believe our positions are wildly divergent, although this may be in part because my own position has significantly developed since the original exchange two years ago.

If I may generalize very broadly, in modern Western culture (probably for ideological reasons), there is at the deepest level a contradiction between concrete specificity and what transcends it. Kant and German Romanticism come immediately to mind. For example, it is said that sociology arose in response to that problem and has been based on ever since: what is the relation of the individual agent and social totalities or institutions? Another example would be the crisis in German historicism in ca. 1900: a methodological obsession with "wie es eigentlich gewesen" threatened to make circumstance determinative of human action rather than moral values.

People have struggled in a multitude of ways within this Gordian knot of mind-body duality. What I have tried to do is to avoid it altogether by redefining action. While action theory tends to see things in terms of the relation of action and intention and takes the physical world as given, what I've tried to do is to define physical world in terms of physical action and in a way that represents it as capable of "self-transcendence". While this is not a new idea (Bohm, Koestler, Bergson, Nagarjuna), these efforts are to varying degrees not ontologically monist, which I assume is absolutely mandatory today. I rely on a modal realism, although not in the now popular fashion of Lewis. I'll be giving a paper in two weeks on the ontology of process that does this. I'll try to get it up before hand on

Given all this, let me address some of your points very briefly.

1. If your point is that Enlightenment thought is too diverse to cram into neat conceptual categories, I'd agree, although we cannot avoid the use of such categories to think and communicate. In any case, my work does not depend on an adequate characterization of Enlightenment views. 

2. No, I try to make consciousness a strongly emergent aspect of matter (my paper ventures to explain strong emergence in entirely naturalistic terms). Rather than a "connection" of entities, I represent them as superposed processes. The result is a rather esoteric argument to show why informed action in the world can be efficacious even though there is no correspondence between that information and the world.

3-4.  I'm sympathetic with Schutz' aim to redefine human relations in more positive terms, but I sharply distance myself from his phenomenology. I read some Zizek and some Schweizer (on Bach) long ago, although I have not read his "Reference for Life". 

5. Yes, memory has to do with the future, but I adopt a deflationary approach to time. Too much to summarize here, but to be cryptic, I take what we call the past to be a mental construction based on the depositing of differentiated persistent structures in episodic memory; the future is the probability distribution of the "present"; since I take a process view, there is no interface between a reified past and future or point in time that is the present.


Linking action to semantic memory
Reply to Haines Brown
Mr. Brown

I feel, it is impossible to make justice to your qualified and inspiring reply in the framework of such forum. Therefore, I look eagerly foreward to your paper.

After your clarification, most of your ideas seem not so far distant from mine. Devil, as usually, lies in the detail. In order to leave you with a short feedback, I will only mention two questions that stroke my mind when reading your comment.

Action and Value
True to the topic of this forum (philosophy of action), action is probably the most challenging point to the reader here. My question is why you separate action from values when you refer to the contreversies in historicism (and Methodenstreit?) and defend "moral values" over "human action". Trained myself in Austrian School, I consider action itself as the expression of a value statement which, in my opinion, is implicit in the concept of intention.

Materialism and Time
Honestly, I doubt that we can have clear ideas of materialism and material without, at least silently, making assumptions on time and the non-material, i.e. mind. I know it is an extended praxis in modern philosophy to search for solutions that avoid apparently unsolvable problems. Language philosophy tries to avoid metaphysics alltogether. The cost of this is that we obscure our metaphysical position, which makes the result not more true or clearer.

Time seems to be the point where we two most oppose our ideas.

Some arguments on time...

A) How can we make "probability distributions of the present"?

Present is immanent. as soon as we abstract it and deduce probabilities we are speaking of the past, aren't we?

B) Further, do I understand you right, that the future is an extrapolation of a reified past in a probabilistic way?

Speaking of probability, we have to distinguish between class and case probability. Class probability is used in natural science (e.g. predicting weather from a set of data of past occurencies) and also in many schools of sociology (e.g. prediction of elections). In the field history and economics (especially concerning leadership and business strategies) the true interest lies in the explanatory power of case probability and the human ability to imagine a future beyond predictions of class probability (as Mehmet II when he decided in 1453 to transport warships overland and conquered Constantinople!)

C) What do you think about my idea, that there is always an element of creativity in the construction of both future and present Reification of the past, as I think, means that the past is not only a simple reconstrutction of deductions from material objects but also a creative actualization of it, always charged with values.

A framework for discussion
I'm not so familiar with Nagarjuna, who, as far as I know, belongs to the school fo Mahayana Buddhism close to Platonian ideas of time. I feel that in Asian philosophy Theravada Buddhism, or the Chinese Huayan School (actually belonging to Mahayana) or Bushido are much closer to my philosophy. We find similar ideas in European thinking by Heraclitus and in more recent time by Hegel and also Bergson.

Bergson seems to be a common denominator of our two lines of thinking. I will try to find answer from within his theory in order support my arguments, this may make discussions easier.

I hope to read your paper soon and will try to honor it with more concluding statemetns.

Looking forward to your thoughts.
Best regards.
Tabea Hirzel

Linking action to semantic memory
Reply to Tabea Hirzel

You said, our tho

> Action and Value

> True to the topic of this forum (philosophy of action), action is probably the most challenging
> point to the reader here. My question is why you separate action from values when you refer
> to the contreversies in historicism (and Methodenstreit?) and defend "moral values" over
> "human action". Trained myself in Austrian School, I consider action itself as the expression
> of a value statement which, in my opinion, is implicit in the concept of intention.

I'll find it much easier to reply in relation to a paper I just finished than in terms of the thread. My apologies if there's a clash.

To start with, when I say "action", I mean physical action, not the action theory of the social sciences. Action theory has its appeal: conscious intention can transcend circumstances, and it engages values (as in the Methodenstreit), etc. However, I venture to explain both weak and strong emergence of non-probable outcomes simply on the basis of physical action. If I succeed, it accommodates human intentional action, but without a mind-body dualism.

> A) How can we make "probability distributions of the present"?

I'm not sure what the issue here is. I assume that a probability distribution has no temporal dimension. Whether past, present and future are only artifacts of consciousness is another issue. I touch upon it in my paper, but do not pursue it because I define "process" without the causal spatiotemporal frame that is presumed in folk psychology. That is, I don't need to defend a deflationary theory of time and so remain agnostic about it.

>  B) Further, do I understand you right, that the future is an extrapolation of a reified past in
> a probabilistic way?

No, not an extrapolation or trajectory into the future. Reification comes up when we speak of human consciousness, but my point is instead based on physical action. The argument is too outrageous to summarize here, but basically I define actuality as the most probable accessible possibilities in a probability distribution, and it  constrains the distribution by grounding it, and the resulting probability distribution is simply what we think of as the future.

> C) What do you think about my idea, that there is always an element of creativity in the
> construction of both future and present Reification of the past, as I think, means that the
> past is not only a simple reconstrutction of deductions from material objects but also a
> creative actualization of it, always charged with values.

Here we're closer. One difference is that I see nature as "creative". I'm not being objectively idealist here. I argue that in naturalistic terms all processes are an actualization of real possibilities, and this probability distribution is the future, which does not reduce to actuality, to a so-called initial state or base level.

A framework for discussion
> I'm not so familiar with Nagarjuna, who, as far as I know, belongs to the school fo Mahayana
> Buddhism close to Platonian ideas of time.

It seems to me this is entirely possible. In my paper I address this. I believe the basic issue is whether possibility is open or grounded. Different interpretations of the Buddhist or Daoist traditions disagree over this point. I was being generous to Nagarjuna. In either case, I don't see how it could be understood as implicitly Platonic, which is ontologically dualist if it reifies possibility.

Bergson also seems to be ambivalent. He surely does not deserve all the abuse heaped upon him by Anglo-Americans. I suspect the basic issue here is "force", which he takes for granted. In the natural sciences there's a growing sense that there is no such thing as "force" as a primitive. All Bergson does is to argue that nature can be strongly emergent. He offers an acceptable description, but in contemporary terms does not really explain. 

I'll be putting my paper on line shortly, perhaps both and philpapers and my own website. It touches on all the issues you raise.