[1.0] A thought is (i) something that exists in the mind while the mind is conscious of it or (ii) it is something that is merely capable of being brought into consciousness (it can possibly be thought of) from (a) memory or (b) the mind's internal or external environment. (b) does not make any metaphysical assumptions, as they are not necessary for such investigations into the nature of thought, but leaves all possibilites of the metaphysical situation open.
[1.1] More on (ii): A thought need not be a conscious thought as most thoughts stored in the normal adult human's mind are not capable of being made aware all at once. This is because the amount of thoughts available to a normal adult human mind would overwhelm the individual's consciousness and not serve to help navigate thoughts as in thinking about something or thinking from one thought to the next.
[2.0] All possible thoughts are possibly available and this means that there are infinite thoughts possibly available to a sufficient mind in a world that is infinite- such as the world of the mind as opposed to empirical reality. By sufficient, I mean a mind sufficient enough to represent the full spectrum of infinite thoughts possible.
[3.0] A thought is synonymous with a perception, a mental representation, a mental image, an idea, a conception, a notion, or a recollection. All of these terms mean or point to, the same thing in reality, whatever reality may in fact, be, should there be a fact concerning the answer to such a question. Some of these thoughts, like a perception, may be simple thoughts that have no conceptual form. The simple perception of the sound of thunder is self-contained an need not point to any other parts.
[3.1] Colors, pains, smells, and tastes can all be seen as examples of simple thoughts. A mental representation can be simple or complex. A complex representation could be that of a car, a horse, a planet, or a spaceship. All of these complex representations require form, color, causal relations, and other simple thoughts to combine and form them.
[3.2] We could imagine only being focused on the word "blue", as it appears here on this page in white, but not, as the word usually does, trigger us to think about the perception of blue. The perception of blue and the perception of the white word "blue" are different. However, under normal circumstances, the word relates to the perception but the perception need not relate as strongly, at least, to the word.
[3.3] One can imagine blue and all the things that are blue without necessarily accessing the word and this would be even easier for someone who has never heard of any color referent but has seen many blue objects. The blue water may remind someone of the blue vase and the blue car. But, never, does the word "blue" need be part of any of it. This would be likened to association of thoughts or ideas or representations based on a common property, namely the blue content of each.
[4.0] A thought is not the same as a concept but is related to a concept in that the thought will feature parts or wholes of the concept of the thing in question.
[4.1] For example, a thought of an elephant may be the static image of an elephant like the one to the left of this page, but it does not necessarily include other facets of the concept that include myriad related facts one believes about elephants, such as their social structure, their movement patterns, or their noisy trunks.
[5.0] A thought is capable of being stored and/or accessed in memory. Memory could mean the brain, the unconscious mind, or the totality of existence and reality itself.
[5.1] A thought, then, need not be a representation that is currently in a mind's consciousness but only be capable of being brought into a mind's consciousness as in the case of, for example, your stored memory of the first time you drove a car or your memory of cotton candy.
[6.0] A thought that is currently being represented in one's consciousness has precedence over all other potential thoughts and is quickly, while one thinks, replaced by thoughts that are related to that thought or some other stimuli in the internal or external environment of a mind.
[6.1] Perhaps, for example, you are thinking of someone who you knew long ago but have not seen in many years when an earthquake happens and refocuses your mind to the shaking objects and furniture and buildings around you.
[6.2] The thought of the person you knew long ago is not lost in this case, but put back into potential consciousness in the form of memory. Perhaps, once the shaking ceases and all returns to normal, you decide to return to the thought of the person you once knew. This thought is not just like a photograph- though it can take on the static characteristic of a photograph.
[6.21] Digital images are more likely the way in which the mind, given certain physicalist assumptions, stores imagistic information given the array of neurons in the eyes can be said to have a finite X and Y or horizontal and vertical, "resolution" as on a computer.
[6.3] Rather, it is more like a mental movie complete with sensory memory and multi-faceted and multi-dimensional aspects. You may remember how this person smiled, they're accent in speaking, the way they wore their hair, and the way they walked, for instance.
[6.4] All of these disparate portions of the concept of this person can be reloaded into memory at will unless that memory is blocked or lost for some reason that could have myriad causes- such as brain damage or alzheimers or a simple lack of connectivity between that memory and one's conscious thoughts.
[6.5] The forgetting of a name to a face, for example, in the case of an acquaintance or a famous actress, is precisely the temporal delay between one portion of a concept and another.
[6.6] The face is part of the concept of this person and their name is another though it is not always easy to quickly and perfectly travel between a name and a face and vice-versa, depending on the mind in question.
[6.7] Often, rarely connected representations, even about the same individual, can be difficult or time consuming to connect in conscious thought because the search routines have not formed a strong link as in the case of the name and face of one's spouse, for example. However, in cases like alzheimers, even the most solid of mental representation linkages or strong connections can be damaged or severed completely.
[7.0] Thinking is a process of acting on thoughts or moving through concepts to new thoughts.
[7.1] Thinking about rotating a red square, for instance, involves (i) having the simple idea of red in memory that is referenced by the word "red", as well as (ii) having the concept of a square that is referenced by the word "square", (iii) having the ability to form a "red square" and then (iv) having the concept of "rotation" as defined as moving in a circular motion in one direction around an axis and having the concepts of (v) circular, (vi) direction, and (vii) axis. This is not necessarily exhaustive of all the thoughts, either simple or complex, that may be needed to rotate a red square in one's mental space.
[8.0] A thought process refers to the associations between concepts and the ability to use logic and imagination to manipulate or examine the causal relations of various representations.
[9.0] I have an essay examining the theories of Donald Davidson and the relationship between language and thought. This essay poses the question, "Can there be thought without language? And, if so, what kind of thoughts can be had by a non-linguisitic or proto-linguistic being or mind?"
Review of Donald Davidson's Core Ideas:
_______________________________________________
(a) A theory of interpretation can not be based on words/gestures apart from beliefs
(b) A theory of interpretation can not be based on the intentions of the speaker (Grice)
For Davidson, (a) is the case because we can not base meaning upon something independent of the speaker's beliefs. If we base meaning on convention of time and place, then we can only assume we have gotten at the meaning of some utterance if the speaker has a whole set of beliefs about the utterance that correspond to the conventional meaning of the utterance in a given time and place. But, Davidson thinks that the notion of 'convention' is not well definable, and is not philosophically interesting.
For Davidson, (b) is the case because there is an unresolvable problem in trying to describe or define detailed information about a speaker's intentional states (beliefs, desires, etc) without using linguistic terms. The intentional states, asserts Davidson, must themselves be non-linguistic if thought is to be prior to language. But, in what way can an interpreter interpret the intentions of a speaker without describing them in a language?
There is, then, a problem of an interpretation circle. We can not base interpretation on the utterance itself or on the intentions of the speaker. We must have one to descipher the other, and we can gain knowledge of neither independently of the other. The solution, for Davidson is a truth-conditional theory of meaning.
Davidson, to solve this interpretation circle must assume that:
-
Agents generally utter things they take to be true (principle of charity)
-
Agents generally act to maximize utility (principle of rationality)
If we examine the truth-value of sentences at some time, in the context of some external circumstances, and accepting (1) and (2), we can determine the meaning of a sentence.
Truth-Conditional Theory of Meaning: an utterance X means the conditions or state of affairs that, if realized, would make X true. (How would X be true? If the language community at the time and place of X agreed that X is true.)
To have thoughts, a creature must be an interpreter of the language of others. This is because, in the absence of language, a creature can not attribute fine-grained thoughts on the basis of action alone.
Also for Davidson, a creature can not have a belief unless it grasps the concepts of truth and falsity, which arise only in the context of interpretation. Using the truth-conditional theory of meaning, creatures can grasp the difference between true belief and false belief, thereby coming to have beliefs.
Summary of Davidson's View of Thought/Language Relationship:
_____________________________________________________________________
For Davidson, a thought is a fine-grained mental state whose content is as distinct as the concepts picked out by symbols in a natural language. For example, if an interpreter is to attribute a belief to a person based on verbal data, such as some speaker uttering "I believe that snow is white", then the attribution of belief to this speaker (by an interpreter) must have some resemblance to the utterance (or gesture).
To have the thought(s) that "She believes that snow is white" requires the interpreter to possess a number of distinct and gramatically ordered representations which are fine-grained enough to interpret the semantic referents of 'she' (other distinct individual), 'believes' (has an intentional state), 'snow' (a type of thing), 'is' (property-of or equality relation), 'white' (a description of the thing).
Although there is a great amount written in response to many aspects of Davidson’s theory of meaning, I hope to, in what follows, concentrate most on the aspects that are most relevant to the language/thought topic.
Two Main Criticisms of Davidson's Theory Concerning Thought/Language Relationship:
_____________________________________________________________________________________________
Criticism (1) Davidson's view is strictly verificationist in nature: Davidson is basing a theory about the thoughts of a creature, which are unobservable features of some cognitive system, completely based on observable data. The observable data in question includes the speech actions and behavioral actions of creatures. According to Davidson, we have no way of verifying the distinctiveness of the thoughts of a creature based solely on non-linguistic behavior (a dog chasing a squirrel).
But, some critics, such as Carruthers, want to point out that there is a deep problem in the kind of assertions that Davidson is making. Carruthers writes, "…epistemology is one thing, metaphysics is another…That we lack sufficient evidence to ascribe any particular fine-grained beliefs and desires in the absence of linguistic behavior is one thing, whether thought itself can exist in the absence of language is quite another"(foot: Lang, Tho, Consci, p. 24).
For example, if the Turing Test were to be passed, based on Davidsonian criteria, we would have to ascribe fine-grained thoughts to the computer/software that passed the test.
I think that criticism (1), while a valid criticism, does not have much force. If a computer was able to pass the Turing Test, then it would seem as though the computer must be manipulating and referencing rather fine-grained representations, which could be called thoughts. The intuitive force of the Turing Test objection, I think, is to have the reader imagine a computer having thoughts while a baby or most animals does not. This, however, seems an entirely possible scenario.
Also, in defense of Davidson against criticism (1), we do not have any other reason to ascribe fine-grained intentional states to a creature other than in light of linguistic behavior. And, if the creature in question did have (in a metaphysical sense) fine-grained thoughts, then we would (a) have no reason to suppose it did, (b) have no idea why it would have such distinct intentional states, and (c) we would have no idea of how it came to possess such fine-grained intentional states. This is assuming that we hold natural language to reference more specific semantic contents than creatures need to reference in trying to survive in a non-linguistic community.
What if non-linguistic creatures did have fine-grained thoughts?
_______________________________________________________________________
Critcism (1) does not make any positive assertions about the graininess of non-lingusitic creatures. However, the only way in which non-linguistic creatures could come to possess natural-language-like intentional states would be if fine-grained thoughts are innately specified (hard-wired) and cognitive development is not influenced in any significant way by the interpretation of or use of a natural language.
This case would be an example of the most extreme version of the communicative coding view of natural language. According to this view, on a non-charitable interpretation of an early Gricean model, an array of fine-grained thoughts exist prior to and independently of use of a natural language. Natural language, under this view, is used to express pre-linguistic thoughts. But, this view, I think, seems quite improbable under conceptual analysis.
To show problems with this aforementioned model, I will outline an example using a Gricean m-intention model: (i) baby intends mother to believe that he is hungry, (ii) mother recognizes baby's intention, and (iii) mother comes to believe that baby is hungry partly on the basis of (ii). In this case, the baby may utter "hungy" (or something like this). Under an early Gricean interpretation, this utterance may reflect the best semantic referent for the baby's psychological state that could reflect a number of fine-grained thoughts such as "I am hungry for warm mashed peas and carrots".
But, there is a deep problem in attributing any fine-grained intentional states to the infant before the infant has acquired semantic referents for such objects, states of objects, and self-representations. If our description of the infant's thoughts is correct, then the thoughts must exist as non-semantic intentional states and labeling such thoughts with terms such as "mashed" and "peas" misses the actual content of the intentional states. One of Davidson's main points is that we have no way of describing the intentional states of the infant.
It is at least highly conceptually difficult to imagine how non-semantic psychological states could pick out distinct objects, properties, and other representations without having semantic referents to refine, order, and express such intentional states.
It seems completely absurd that a non-linguistic creature could entertain the following two propositional attitudes about the same object: (1) "Venus has set" (2) "The evening star has set". The question is, what does this entail about the nature of languageless thoughts? This brings us to the second criticism, one which I find highly relevant to and weaking of Davidson's overall picture of language as it relates to thought.
Criticism (2) Davidson's definition of a 'thought' is overly restrictive: It is one thing to say that it is necessary to be an interpreter in a language community to have a certain type of thought (namely, a fine-grained thought), but it is quite another to say that it is necessary to be an interpreter of language to have any thoughts at all. It does not seem intuitively correct that thoughts could suddenly appear with language.
Rather, it makes more sense that there are some types of intentional states that may be augmented (to some degree) by language, and that certain types of cognition are, to some degree, influenced by the use and interpretation of a natural language. We may allow Davidson to hold his restrictive definition if, for him, thought means only fine-grained thoughts. Perhaps, other types of thoughts do not qualify as thoughts, because they are not distinctively human thoughts. But again, this seems to restrict the term 'thought' beyond usefulness.
It seems introspectively and psychologically obvious that language plays a role in conscious thoughts that are about other thoughts "I realize that I desire x, but x is bad for me, etc". But, what about unconscious thoughts? These thouhts must have content, and (arguably) have fine-grained content.
Our unconscious thoughts are those that relay information in spatial reasoning, motor skills, and other non-verbal cognitive domains. Davidson may have been concentrating just on the conscious, or distinctively human, thoughts. If this is what he meant, then yes, we do need language to have a thought. But, there are a few other problems with Davidson's characterization of 'Thoughtless Brutes' (animals and infants).
Some Other Thoughts about Thoughts and Language:
_____________________________________________________________
Animals: Davidson’s notion of requiring that to have a belief, a system must possess the concept of belief (which requires language) is not without contest. Fred Dretske has maintained that behavioral data about many animals does constitute reason to ascribe to them certain basic intentional states such as beliefs (1999, ‘Machines, Plants, and Animals’ Erkenntnis 51, 19-31). READ PASSAGE FROM STEPHAN (pg 84 in article).
The philosophical reasoning behind ethological ascriptions of intentional states has come from a number of people, notably Hans-Johann Glock has given a number of arguments concerning the requirements for belief, and the best way in which we should use the term as it is ascribed to the mental lives of non-linguistic creatures (2000, ‘Animals, Thought, Concepts’ Synthese, 123, 35-64). READ PASSAGE FROM GLOCK (pg 56 in article).
Human Language Deficit Case Examples: Russian psychologist (Luria, 1959) separated two identical twins and exposed one to language and the other not. Without language, many cognitive deficits were found in the languageles child after only a period of six months. The case of Genie (Curtiss, 1977) points to cognitive deficits without language use/interpretation/acquisition.
Conclusions:
____________________
I take it to be plausible that language interpretation is necessary for the possession of fine-grained intentional states like those we find in humans because (i) we have no data upon which to attribute such states to non-language users, (ii) it is conceptually impossible to identify the contents of non-semantic psychological states that are (in Gricean terms) ‘hazy and nebulous’, and (iii) we can attribute no reason why or how non-language using creatures would come to possess such discrete intentional states.
However, I find that a slightly deflationary interpretation of what it is to be a thought helps to explain exactly what it is that language acts upon in creating fine-grained intentional states. Perhaps these pre-linguistic intentional states, which are essential to an organisms survival, are refined and discrimated upon by possession/use of a natural language. Thus my concluding statement is: thought is prior to language, but such thoughts can not qualify as distinctly human thoughts. Therefore, language is necessary for our type(the human type) of thought.
From 1999 | Thought seems to be an event:
_______________________________________________
In humans, it is characterized by the presence
of an idea in the mind. If we can perceive of something in our minds, anything at all,
then it follows that we are thinking about that thing. Again, in humans, the brain seems
to be that main facilitator of thought, or the substance in which it occurs.
But this is
not the whole picture. If thought is an event, then what other things have the capacity to
think? Isn't the logic board or processor of a computer also, like the brain, a
facilitator of thought in the sense that it is an event comparable to the electro-chemical
events in the brain that allow us to think? At the most basic level,
thought can be defined as anything that has motion in order to reach an idea, conclusion,
calculation, or any other such idea. Therefore, in this primitive sense of the concept of
thought, even a wooden abacus has the capacity to "think".
Can a computer think? If we mean, "Think exactly like a human" then
the answer is no, because only a human can think exactly like a human. If we mean,
"Think like a human, or think in such a way as to simulate human thought" then
the answer seems to be yes. This is because of the capacity for computers to have many of
the same sensory perceptions that we are capable of, and use them in such a way as to
simulate human thought.
Though the standard computer of today uses a fixed hardware
system, technology can and is being developed to make facilitators of thought (like our
brain) that are capable of dynamic restructuring in order to think more like a human.
The
facilitator of thought in the conventional computer is the processor, and in the processor there is the event of electricity that allows the computer to process
calculations. In one sense, the thinking power of modern computers in terms of pure
calculation is better than that of humans. However, we created computers, and the medium
in which they think. They are, in a sense, still primitive.
Can an abacus think? If the definition of thought is the movement of something
in order to reach an idea, calculation, or conclusion, whether it be in a machine using
tangible parts, or the movement of electricity through a processor, or the movement of
electricity and chemicals, we must then say that an abacus is capable of thought in a
sense.
The event is the movement of the parts, the result is a type of thought, or the
result of the movement. Now, someone has to interpret the result, but do
we not have to interpret the results of all movements that we call thoughts? If we were
not actively interpreting the motion that we call thought in our own minds, what would we
ever think?
The abacus example has some problems on the surface, but once examined
carefully has no distinction from the thought that occurs in other facilitators. We think
internally, through the use on energy consumed in the food we eat and the air that we
breath.
A computer also thinks internally, using the electricity that we provide it. In
both cases, energy must be provided to some form of matter in order for
motion, and thought. In the abacus, the hand of someone must provide the energy to the
physical matter of the structure of the abacus. Is it then, that thought is no different
that movement of matter, which requires energy in one form or another?
Does the ability to think have anything to do with the complexity and organization
of the subject? Think of the abacus, and imagine a substance composed of millions of
minute abacus's that all function together in order to reach more complex results. We
could compare this to the modern computer, because they use gateways to determine simple
calculations based on binary digits.
Each of these is comparable to an abacus in an
abstract sense to show how something very simple can be complex when
compounded. One calculator can do one calculation at one point in time. Two calculators
can do two calculations, and so on. So, if we somehow have a million calculators.
Are there different levels of thought? It seems as though the complexity and
organization of the processes, both the matter which facilitates and the energy that is
provided, determine some level, greater or lesser of thought. A conventional computer may
be said to think more perfectly in relation to pure calculation, but can not think in the
way that we can.
This is because of the composition of the computer and because of the way
in which it is supplied energy. The pathways of a computer are set, and can only perform a
few simple functions in order to reach a conclusion. Our brains have pathways that are
more dynamic, and have more methods of obtaining and utilizing energy. We use both
chemicals and electricity in ways that are so complex and organized that we have yet to
truly understand the workings of the brain.
Thoughts from earlier, 2001:
_______________________________________
The following will be my views on the potential thoughts of beings with no symbol use or interpretation whatsoever. I think that Davidson is correct in pointing out the descriptive failure of our attributions of intentional psychological states to non-linguistic creatures on the basis of non-linguistic behavior alone.
However, this has not convinced me that such beings do not have thoughts at all. Our inability to describe such thoughts, I think, is just the inability for linguistic terminology to capture other non-linguistic forms of representational thought. This gap between our description of such thoughts and the actual content of such thoughts that we are attempting to describe need not constitute a total rejection of the existence of thoughts in some beings.
If there were no behavior at all on which to base ascriptions of thought, then I would agree that there is no basis for ascriptions of thought. However, I find that many examples of behavior in totally non-linguistic creatures constitutes the attribution of some, abeit very simple, thought.
To have the most simple and basic form of thought, I propose, a creature must be able to represent some part of the world in some way that is perceptually available to the creature and in which the representation may be mistaken. This, I think, is criteria under which a a purely behavioral disposition becomes a thought. Such a simple and basic thought is holodoxastic in nature (Glock, 2000).
The holodoxastic thought must be represented in some sensory modality or modalities, such as the visual, auditory, or somatosensory modalities, and must be capable of being mistaken. Upon what behavioral evidence could we attribute a holodoxastic thought? I will attempt a generic example that could be applied to ethological evidence.
Imagine that some creature C tracks or chases some object O (animate or not) until O becomes unavailable to the perceptual mechanisms of C. If C still reacts as if O exists, then I would say that it is appropriate to attribute C with at least some holoxastic thought about O. The problem lies in our saying something like "C believes that O is behind the fence, still exists somehow, or has gone underground ".
We have no way in which to accurately describe the holodoxastic belief in such a way that is evidentially satisfactory. A solution to this which I think is notable is Achim Stephan's suggestion that we say that such beliefs are intensionless (1999). In other words, although the extension of the thought (that which it is about) is constant, we have no way in which to make sense of the intension of the thought, or the way in which the thought is about the extension.
A holodoxastic thought is a thought that is referentially transparent, intensionless, and requires no further thoughts about the content of the holodoxastic thought because there would be no 'parts' of the thought for other thoughts to be about. Our problem in attributing holodoxastic thoughts under some description strikes me as similar to the problem of describing a picture with only linguistic terms- the description of any pictoral image could be made under countless descriptions that would never quite 'capture' the content of the picture.
And, if the holodoxastic thought of some creature is represented in some, foreign to humans, sensory modality such as sonar (bats), electromagnetic sense (sharks) or a somatosensory mechanism (dogs), we would have a very difficult time producing any description that would approach the content of such thoughts.
It should be noted that a holodoxastic thought is the most complex type of thought that I would attribute to a being with no language at all. This is because the idea of having a thought about something that is part of a holodoxastic thought requires the notion of some reference to the part of the holodoxastic thought. Without any symbol there seems to be no way in which the creature could refer to something in their thought.
(b) Thought with Some Non-Natural Language:
__________________________________________________
Once creatures have begun to produce and interpret some symbols, their capacity for thought, I think, increases. The range of languages between the absence of language and a full-fledged natural language is potentially quite extensive. We could imagine that a creature without any syntactical rules for the combination of symbols could have a symbolic lexicon of dozens or hundreds of symbols, a lexicon that would only be constrained by the creature's ability to produce and interpret distinct symbols and its linguistic memory.
But just being able to make symbols does not make the symbols part of a language. In order for a symbol S to be part of a language, rather than just a behavioral reaction, S must be capable of being interpreted by other members of the language community. S, then, means that which it stands for to the interpreters. Once a creature has the ability to use symbols to communicate information about their environment, I argue, the creature may be able to form rudimentary concepts or more than one belief about something.
For example, imagine that an individual creature C, where C is part of some community M of creatures of the same type as C, produces some symbol S which stands for some object O (animate or not) in its environment. Also imagine that object O has some behavioral significance R to inviduals of M to which C belongs. Now, if O has some behavioral signifigance R to the individual members of M, O would have R even if S had never developed as a symbol for O.
But, because S did develop as a symbol for O, there is room for the most simple of concepts. Why? Because S stands for O and O stands for S, where both S and O cause R. How could we test for this most simple of conceptual schemes? If C is capable of producing S in the absence of O in order to produce R in individual members of M, then S can be attributed with the most basic concept of O.
Because this situation is a bit complex, I will use some empirical ethological evidence from studies of vervet monkeys to support my generic criteria for behavioral evidence upon which to attribute simple concepts. Individual vervets (C) have, according to Seyfarth (1980) been able to use (among 5 known calls) the eagle-call alarm (S) in the absence of an eagle (O) in order to produce a 'selfish' behavioral response (R) in other vervet monkeys (M).
Also, members of the vervet community will often check to see whether or not the call is true or false. This, I think, is a plausible example of having the concept of the symbol. A human (C) who knows only that the word 'fire' is a symbol (S) for some thing (O) that also causes a behavioral response (R) in humans (M) can be said to have the most simple of concepts of fire.
The complexity and degree to which thoughts can be influenced by the possession of a non-natural language depends on a creature's (i) discriminatory perceptual abilities, (ii) active cognitive faculties, (iii) symbol producing abilities, and (iv) linguistic memory capacities. (i)-(iv) is only a rough description of the things that go into deciding how much a non-natural language can affect thought. But, non-natural language with some cognitive capacities for processing and arranging symbols can quickly express quite complex thoughts.
Consider some community of fictional super-vervets that live in trees and have five symbols for the following things: (f) food, (d) dangerous, (s) sky-location, (g) ground-location, and (t) tree-location. The possible two-place combinations, if they are able to represent such combinations are (f)(s), (f)(g), (f)(t), (d)(s), (d)(g), (d)(t), and the reverse of each of these, which makes six distinct meanings and two ways of expressing each of the six distinct meanings.
But, if the super-vervets were a larger, predatory species able to, in groups, kill things that could also kill them, and they had the cognitive capacity to understand or represent three-place combinations, they could produce (d)(g)(f), where some combination of these symbols means "dangerous-thing down/ground-location food-thing" may represent a snake, which could be both a predator and prey for the creatures.
Three-place combinations would yield three more distinct meanings with nine possible ways of expressing each of the three distinct meanings. Also, we could imagine they come to understand (f)(d) as a symbol for poisonous foods, which is one more meaning with two ways of being expressed. Together, these five symbols minimally could yield ten meanings.
We do not see very complex use of non-natural languages except in some limited cases of higher primates being raised in captivity, such as the case of Kanzi. And, while Kanzi does exhibit the capacity to combine some symbols in order to manufacture new meanings from them, he does not seem to do so in a way that utilizes the potential complexity of the many symbols he has learned (Savage-Rumbaugh 1980).
Of the criteria I outlined above for the degree to which a non-natural language influences thought, it seems as though Kanzi is certainly lacking in his (ii) active cognitive faculties, (iii) his symbol producing abilities, and very likely his (iv) linguistic memory capacities.
(c) Thought With Natural Language:
_________________________________________
If my account in (a) and (b) is on the right track, then natural language will be an extension of non-natural language that includes syntax, and natural language, or the cognitive abilities that accompany it, will allow for the types of thoughts that normal adult humans can have.
The question is a question of how big the gap is between creatures with a non-natural language and a natural language, where the fundamental difference between the two is that the latter has a syntax or grammatical structure. I think the best way to answer this question is to compare two languages, each with the same number of semantic referents but in which only one has any grammar, and see the different possibilities.
If one recalls my description of the five symbol lexicon of some ficticious super-vervets, then we can see that they at least have five meanings that they can communicate. Remember that possible combinations of the five symbols could potentially yield up to fourty plausible meanings, and an addition of another useful symbol for 'far' could double this number of meanings. However, any use of the five symbols for more than five meanings requires that the super-vervets be able to understand or represent the combinations of meanings.
Remember that if they can represent two-place combinations of (f)(d)(s)(g) and (t) they will get 6 new meanings, if they can represent three-place combinations, they will have 3 new meanings. Also, if they can represent each of the possible combinations of the two-place combinations differently they will get a total of 12 new meanings from them, and if they can represent each of the possible combinations of the three-place combinations differently they will get a total of 27 new meanings from them.
My reasons for restating the example are the following: The ability to represent each of the combinations differently requires an underlying ability to represent the referents of symbols based upon their sequential arrangement. This ability to represent the same symbols differently, depending on the way in which they are arranged, is a sign of grammatical cognition, or cognition that is ripe for the addition of symbols like 'is', 'on', 'and', etc.
If the super-vervets could begin to understand "food-thing ground-thing dangerous-thing" as meaning something different than "dangerous-thing food-thing ground-thing" then their symbolic lexicon would be ripe for something like 'is','on', and 'and' which would allow for the more clear "food-thing is on ground-location and is dangerous-thing" and "dangerous-thing is on ground-location and is food-thing". The former of these more grammatically correct sentences could mean that the thing is primarily food but is also dangerous (such as poisonous fruit), while the latter could mean that the thing is primarily dangerous but also food (such as a snake).
Because of these thoughts, it seems likely that (1) language is largely a reflection of the cognitive abilities of the creatures that use such a language, but (2) language acquisition is, for the most part, necessary for the utilization of lingusitic cognitive abilities (and for the production of complex thoughts), and (3) the ability to use and interpret language was plausibly a driving force behind primate evolution of better cognitive systems able to produce language.
Therefore, I propose that natural language requires a natural language cogntive base (NLCB), and the utilization of a NLCB requires the input of a natural language (within a certain widow of availability)
With a NLCB and a natural language, humans can form self-representational thoughts, they can 'entertain' thoughts, they can voluntarily manipulate the content of their thoughts and they can form multiply embedded thoughts about their thoughts. Natural language, I do not think, can not be used without a NLCB. Deprived of language, a NLCB may still display some complex cognitive ability above other creatures, but seems to fall well below the mark of normal ability with the aquisition of a natural language.
I find the most plausible account of the relation between thought and language, which is in agreement with my own thoughts on the issue (from question is Juan-Carlos Gómez's accountwhich supports and is complementary to both a Fodorian and Gricean position. One of Gomez's main points is that:
...ostensive communication, with its openness and creativity, may have been an important factor in 'calling for' the evolution of linguistic grammars and meta-representational theories of mind... A plausible first step was the combination of ostensive and semantic systems... [which] created the cognitive conditions that made specialized devices such as grammars and meta-representations useful (Gómez 1998).
Gómez's account, which I think is complementary to my own account, is supportive of a generally Fodorian position on how a pre-linguistic cognitive structure for processing symbols is crucial to formation of complex thought. This is similar to what I have called a natural language cognitive base that must be in place in order to process symbols acquired through gaining a natural language. Also, Gómez's account supports a Gricean interpretation of how language is used to communicate the intentions of speakers, intentions which often have non-linguistic content.
That thoughts can exist independently of language is supportive of a Gricean interpretation and an overall communicative conception of language. However, I would call my overall account a kind of deflation of Grice's account and a middle ground between the communicative and cognitive accounts of language. This is because my account recognizes that language must also be acquired in order to use pre-linguistic cognitive capacities and therefore such cognitive capacities are dependent upon acquisition of a natural language.
b) I find the most implausible account of the relationship between language and thought to be any account that thinks that thought is completely independent of language or any account that takes thought to be completely dependent on language. So, a very strong Gricean interpretation, some interpretations of Dennett, and the view of Davidson would all be examples of understatements or overstatements of the degree to which I think language affects thought.
Although, I do feel a certain sympathy for the cognitivist positions in that they generally define thought as the type of thought that would be possible only with a natural language. However, I ultimately feel that these strong cognitivist conceptions of language's influence on thought are overly restrictive. And, I am certainly opposed to any who would argue that thoughts are completely independent of language and that language's only purpose and function is within the context of communication as the translation of thoughts into a language.
References:
_________________
Glock, H.J. 2000. "Animal Thoughts and Concepts," in Synthese, 123: 35-64.
Savage-Rumbaugh, S. 1980. "Do Apes Use Language?" American Scientist, 68(1): 49-61.
Seyfarth R.M., Cheney D., and Marler, P. 1980. "Vervet Monkey Alarm Calls: Semantic Communication in a Free-Ranging Primate," in Animal Behavior, 28: p. 1070-1094.
Stephan, A. 1999. "Are Animals Capable of Concepts?" in Erktennis, 51: 79-92.
