Reflection on the content and the reference of the Idea of Representation
Review of Vacariu et al.'s Toward a very idea of representation
'the debates in Cognitive Science focus on the problem of representation. Unfortunately, these debates regard either representational format(as in the computationalism-connectionism debate), or the problem of representational content(as in representationalism-antirepresentationalism debate), but they are not concerning with the nature of representation. Thus one question legitimaly arises: how can beyond of these debates? Our believe is that what completely misses in Cognitive Science is an abstract theory of representation by means of wich it is possible to go beyond of these debates'(Vacariu et al. 2001, 290).
'Authors like Mundy, Swoyer, Suppes, Mormann and Ibara, etc., dealt with a general theory of representation. Thus, it is not about creating a new abstract theory of representation, but only to search wich theory is fitting best with the problem of mental representations.
For instance, the concept of Swoyer's structural representation (an isomophism from relational system A regarded as a set of the physical objects to the relational system B regarded as a set of the numerical entities) grasps the aspects of symbolic representations'(291)
'The general theory of representation releases by Mundy is based on structured and unstructured representations. The unstructured representations are different to the structured representations in the sense that the elements of set A are mapped to elements of B by a function f wich is understood to represents the numerical relations of the elements of A, but the numerical and physical relations are not explicitly specified (Mundy 1986). We believe that this concept can capture the aspects of connectionst representations wich can not be isolated, each computing unit being involved in representing several things.
In the case of dynamical systems theory maybe it is improper to talk effectively about the notion of representation. But we can identify some structures that are preserved through the continuous changing within the state space.'(Vacariu et al. 2001, 291).
I agreed that the intention of exceed certain disagreements can be intersting and good.
But, Vacariu et al. suggested, by they article title, that the abstract theories that they presented, at least, tend towards the idea of representation.
I consider that the idea of representation is older that the Cognitive Sciences.
In addition, this idea is sometimes improperly used.
I will propose a content as the content of the idea of representation and I will compare the previous three ideas with my idea.
For Edelman, representation is not similar with something from environment, but it present similarities between cognitive or conscious contents/informations.
Clark considered that,'A representation is a certain stored information wich stands-in for something, its function being that of carrying specific information. However, the bear existence of a correlation between a certain inner state and some bodily or outer parameters is not enough to establish the representational status of that inner state, more important being the nature and the complexity of correlation'(Clark 1997) cf.Vacariu et al.2001).
I think that, there must be first order conscious states. Even if perception is not independent of previous experience. At least temporally, first order conscious informations must exist. The conscious first order flow is the fundamental matter of cognition.
There are people that concentrate their atention on the study of human visual system, because they had supposed that cognitive process advance in the folowing way:
vision, perception, idea(cognition).
This implies that:
Idea->perception->vision->photons.
However, I think that the first order conscious material of cognition come from all sensorial systems, not only from visual system.
It is clear that, in the actual scientific cognitive process we does not start only with pure perception, but also with the fundamental theories and ideas wich we gained until now. We do not start again from 0.
The idea of re-presentation is a composed idea:
it suggest an action, representation.
more exactly, the action of presentation something that was already presented before.
Maybe, in some point of the interval of cosmic evolution many events are repeated without of our cognition about them. However, representation
implies an agency. On the other hand, if the quantum events are reversible, as is pretended by some physiciscts, and if we would have the power to
control the behavior of quanta, then there would posible also non-cognitive representations.
That is, the idea of representation can be suitable for non-cognitive physical events.
Now, given this two characteristics of
the very idea of representation we find that:
this idea does not suggest especially a structual or formal equivalence.
what this idea suggest is not primarly a similarity.
this idea does not involve primarly the objective aspect of mental representations.
this idea does not implies only something cognitive.
this idea does not implies especially and directly a complex correlation.
All the very idea of representation implies is a temporal repetition of something wich is identic non-temporaly.
Given the fact that, we can to repeat in some limits, and therefore to represent, some states or actions from the past(e.g., feeding), there are non-cognitive representations. That is why:
Representational(thing)do not implies Neuronal/Neural(thing)
or
R(x)do not->N(x)
Vegetables also feed.
Identity is an idea wich involve indistighuisability.
On the other hand, identity can be absolute or relative.
The idea of representation involve only a relative identity, because suggest a temporal difference.
That is why the idea of representation involve a temporal non-identity.
From a physical point of consideration what is identic in univers is some forms.
For exemple, there are hidrogen atoms but they are not composed by the same physical particles, but only by the same type of physical particles.
That is, some physical processes and some physical structures may have a general-form identity.
From the point of view of thermodinamics the evolution of an isolated system it is produced only in one sense (towards increase of entropy), wich cannot by inversed by itself.
But, at the microphysical level is posible reversibility.
And, at the biological level there are so called 'imortal' cells.
As has been said by Hutanu, 'from the point of view of termodinamics, in the case of a reversible process, the system must to return in the initial state 'by the same way' by wich it passed (by the same transformations) from this state in the final state. In addition, the environmental initial state must be the same'(Hutanu, 1983, 58).
But, I think that, even if the system would can to return in the initial thermodynamical and dynamical state, by the same transformations, this would not make posible a representation of a point of time.
Argument: Every point of time has a place in a temporal succession. The evolution of this succession can be only positive. As in matematics, starting from 11 we cannot advance to 10. Even if two state of the universe would be indistinguishable thermodynamicaly and dynamicaly, however, they cannot share successively the same point of time.
That is why my fundamental idea is that
there cannot be a absolut representation[of a point of time].
When we mentaly represent the same point of time, we are in a diferent physical point of time.
And, if we would represent a physical state in the environment, this state would ocupy a different point in the cognitive succession of time. That is, I want to say that, the absolute succestion of time is like a scalar, is quantitative, additive.
There are people(maybe Stefanescu*) wich consider that the study of visual system is the most important to understanding mental representations.
I think that, at least first order visual representations cannot exist, because:
the absolute first order visual conscious contents are not re-presentations. Brain's representations have as a condition of posibility a recognition of an identic content or the power to store and recall conscious contents.
On the one hand, if the same brain's areas are used in creative imagination, however, this imagination is not pure representational, on the other hand, this imagination is not first order vision.
In addition, there can be non-visual re-presentations.
The conditions of the posibility of the cognitive re-presentations are:
the capacity of having actual conscious contents.
the capacity of store conscious contents.
the capacity of recall the stored or coded conscious contents.
Now, I return to the first ideas selected by Vacariu et al. to see if they fit the very idea of representation.
Swoyer's idea mean an isomorphism between something physical and something matematic(cognitive).
First of all, we should not accept a fundamental-natural diffrence between (visual, auditive, olfactory, taste or epidermal) analysators and what is analysed.
Indeed, what is analysed can be other sensorial analysators.
Both are identic in nature, that is they are physical or material in nature.
And, with regard to isomophism, this idea does not contain the idea of representation, in this case it suggest rather something similar with perception. And Vacariu et al. used the phrase 'perceptual representational constancy'. But, even in case of perception the idea of isomorphism is not always clear implied. For instance, the taste of bitter share the same form or structure with its material stimulus?
Or, better, when we see somthing red, this redness is isomorphic with some external object or with the photonic flow reflected by it?
On the other hand, more important is that the very idea of isomorphism does not implies representations.
Indeed, isomorphism is a relation between some entities or at least temopararly stationary states. But, the very idea of representation, of wich content is older than cognitive science, suggest an action: the action of presentation something that was presented before.
Mundy's idea is based on structured and unstructured representation. The structured representation is similar with the previous case. In the case of unstructured representations we have a function between unspecified physical relations and unspecified numerical(cognitive) relations. Maybe, a function between two unspecified structures. And, this structures thoug that they are unspecified they can be both in the same time. Again, it is suggested a direct, sensorial relation. The action of presentation is not implied at all. It is not specified that one of unstructured states must be presented before. Therefore, this abstract theory is not identic with the very idea of representation. Therefore, it cannot subsume any type of representation.
Finally, we should to consider the idea of cognitive preserved structures. Is this idea equiintensional with the very idea of re-presentation?
The world is changing, that is why the structures of stimuli cannot be preserved absolutely. However, Vacariu et al. discused about an 'interval of similarity of perceptions'. That is, our perceptional system it is not sufficient of fin to detect the microscopic level changes, it detect only the macroscopic ones. But, this is perception not representation. It is posible to be a preservation of structure between what we carried now in present and what has been presented, sensorialy or perceptively, before. But, the previous theory involve also non-representational structural preservation. That is, the structural preservation can be substantiated by the imperfection of perceptual system, it does not contain any reference to the action of recall. That is, the structural preservation is not due a recall of something presented before, but both states are perceptual in different temporal points. This structural preservation can be only due a two slow change of external stimuli, or due randomly, it does not involve necessarily any recalling. That is why this theory is too extensive.
When we recall a film that we saw with certain years before, like Terminator II, we truthly represent something.
However, on the one hand, certain images we had forget, that is, the succession of images is incomplete, on the other hand, our images are incomplete both structurally and qualitativelly. That is, some contents that are about the same stimulus and at least one is a product of a re-presentation, but they are non-identical contents, both structurally and qualitativelly.
That is, there are possible partial or incomplete re-presentations.
Notes
*Stefanescu Adina is a postgraduate wich received a maximum qualificative for its licence work being coordinated by a proffesor of philosophy of mind, but its work was a synthesis of approaches of the visual representations and visual system from the point of view of cognitive neuroscience (cf. its proffesor Paul Kun).
References
Mundy, B.:1986, On the General Theory of Meaningful Representations, Synthese 67, 391-437.
Swoyer, C.: 1991, Structural Representations and Surrogative Reasoning, Synthese 87, 449-508.
Vacariu G., Dalia T., Vacariu M.: 2001, Toward A very Idea of Representation, Synthese, 129, 275-295.
