Language and sentience in whales and dolphins

[jim stone]

But language is not just "vocalization or signs". It's extremely complex--and it can express extremely complex ideas. We don't even know how complex the ideas of dolphins, parrots, and bonobos are. Do they think things like, "I can't believe that numbskull Bob didn't give me a treat after I did all those tricks for him. If he doesn't give me a double ration tomorrow, I'm not going to do any more, no matter how he pleads, but if he does, maybe I'll consider forgiving him. But maybe he didn't give me a treat because Trixie is better at tricks than I am. Wow! Maybe they're planning to get rid of me. I'd better be careful until I know for sure."

I guess I think it's vain and unscientific to automatically assume that complex animals that guard borders by going on patrols, work systematically to hunt small mammals, teach their kids to make simple tools, form alliances that lead to changes in leadership, share food based upon favors done by other chimps (and withhold food from others for behavior not supportive), etc. don't have complex thoughts of their own. Observations by Goodall have documented such behaviors in the wild. Other chimp researchers have observed similar behaviors. If you've never heard Goodall speak, and had the chance to talk with her afterwards, she's either one heck of a liar (but then how do you explain the film footage, consistent observations over 30+ years, and other independent researchers reporting the same behavior elsewhere?) or we need to remove our "special" view of ourselves and realize we're just animals, too.

Granted, I'm talking observable behaviors here, because chimps don't have the vocal chords to support human speech, and their hands and arms really aren't well designed for ASL, either. I can't talk with chimps, but their behaviors show evidence of planning, memory, social relationships, etc, and anything short of measuring observable behaviors in non-speaking organisms is only guessing.

Eric

This is all very interesting. The fact is that we appear to have
a syntactic organ, as Mike put it, not that it's located
anywhere locally in the brain, that enables us to
use grammar to put signs together into new sentences;
we can generate and understand sentences that
have never been expressed before--this one, for instance.

And this seems to be used in ocurrent thoughts,
where we speak to ourselves in real time, as it were.

Well, so far, there is no strong evidence that non-human
animals have this syntactic ability--they may have vocabularies,
like Alex, but they aren't any good at sentences.
They may have rudimentary language, but not
language like ours, that is, syntactical language,
apparently.

At the same time, for all the good reasons you've just
given, animals are capable of representing cognitively
complex states of affairs. They can think and reason
about propositions that we express in sentences but
they can't.

This suggests that a) either they DO have syntactic
abilities, or b) it is possible to represent and reason
about complex states of affairs, propositions, without
a grammar, without syntactic ability.

This gets us to the question of the nature of thought.
I go for b. Plainly Max, my dog, has concepts, quite an
array of them, that he deploys to realize that
I will give him the treat if he goes into the pen,
but it isn't something he thinks occurrently ('Hey, he'll
give me the treat if I run into the pen!') and the
thought doesn't seem to be realized in any syntactic
form.

Jerry Fodor maintains that all thought, including animal
thought, takes place in a language of thought,
which is an unconscious innate syntactic language,
with inate symbols with built in meanings. So anything,
chimps, humans, dogs, that can think and reason
has a syntactic language built in as it were.

The language of thought isn't conscious, it doesn't
confer linguistic abilities like ours, Fodor thinks,
still something like a real full blown language underlies thought
always.

There are some difficulties here, I believe.
First, if all human beings have an innate syntactic
language, one would expect all human languages
to have the same grammar--the grammar of LOT.
It would naturally inform our speech, and it would be
positively counterproductive to speak a language
with a different grammer from LOT, because it
would take time and energy to translate back
and forth from LOT, where the real thinking is
going on.

Second, if non-human animals have a LOT, with
a syntax, given their striking intelligence and
social organization, all that would prevent them
from talking to each other in sentences is
the inability to vocalize sufficiently. They've
got all the cognitive stuff built in. But, given
the extraordinary advantages of such speech
in socially organized communities, it is terribly
hard to believe that evolution wouldn't have
broken through in several places to
full-blown language. Wolves would talk,
chimps would talk, we wouldn't expect
to be the only ones, certainly.

On the other hand, if LOT is mistaken, if animals
have no inate grammar, well, how do they
represent complex states of affairs, as I agree
with you they do?

So we have arrived at the question:

How does thought work? Can something
without a grammar nonetheless deploy
concepts to represent complex states
of affairs? How?

[Walden]

sentences

happy

[jim stone]

In that picture you posted of the kid and the chimp,
which were you? Affectionately

[Walden]

In that picture you posted of the kid and the chimp,
which were you? Affectionately

I am the one farther from the camera. We have several family pictures of me and the chimp, and I've grown up hearing the anecdotes, though I do not remember it, as I was slightly too young.

[Darwin]

But language is not just "vocalization or signs". It's extremely complex--and it can express extremely complex ideas. We don't even know how complex the ideas of dolphins, parrots, and bonobos are. Do they think things like, "I can't believe that numbskull Bob didn't give me a treat after I did all those tricks for him. If he doesn't give me a double ration tomorrow, I'm not going to do any more, no matter how he pleads, but if he does, maybe I'll consider forgiving him. But maybe he didn't give me a treat because Trixie is better at tricks than I am. Wow! Maybe they're planning to get rid of me. I'd better be careful until I know for sure."

I guess I think it's vain and unscientific to automatically assume that complex animals that guard borders by going on patrols, work systematically to hunt small mammals, teach their kids to make simple tools, form alliances that lead to changes in leadership, share food based upon favors done by other chimps (and withhold food from others for behavior not supportive), etc. don't have complex thoughts of their own. Observations by Goodall have documented such behaviors in the wild. Other chimp researchers have observed similar behaviors. If you've never heard Goodall speak, and had the chance to talk with her afterwards, she's either one heck of a liar (but then how do you explain the film footage, consistent observations over 30+ years, and other independent researchers reporting the same behavior elsewhere?) or we need to remove our "special" view of ourselves and realize we're just animals, too.

Granted, I'm talking observable behaviors here, because chimps don't have the vocal chords to support human speech, and their hands and arms really aren't well designed for ASL, either. I can't talk with chimps, but their behaviors show evidence of planning, memory, social relationships, etc, and anything short of measuring observable behaviors in non-speaking organisms is only guessing.

Eric

This is all very interesting. The fact is that we appear to have a syntactic organ, as Mike put it, not that it's located anywhere locally in the brain, that enables us to use grammar to put signs together into new sentences; we can generate and understand sentences that have never been expressed before--this one, for instance.

And this seems to be used in ocurrent thoughts, where we speak to ourselves in real time, as it were.

Well, so far, there is no strong evidence that non-human animals have this syntactic ability--they may have vocabularies, like Alex, but they aren't any good at sentences. They may have rudimentary language, but not language like ours, that is, syntactical language, apparently.

It seems unreasonable to assume that animals that don't have the capability to speak or sign would evolve the complex mental structures needed for communicative language. What would the payoff be for the resources required for such a thing?

At the same time, for all the good reasons you've just given, animals are capable of representing cognitively complex states of affairs. They can think and reason about propositions that we express in sentences but they can't.

This suggests that a) either they DO have syntactic abilities, or b) it is possible to represent and reason about complex states of affairs, propositions, without a grammar, without syntactic ability.

This gets us to the question of the nature of thought. I go for b. Plainly Max, my dog, has concepts, quite an array of them, that he deploys to realize that I will give him the treat if he goes into the pen, but it isn't something he thinks occurrently ('Hey, he'll give me the treat if I run into the pen!') and the thought doesn't seem to be realized in any syntactic form.

A fair amount of information processing (is that "thought"?) takes place unconsciously. When a kid heads for the spot where a fly ball is going to come down, she probably doesn't do a lot of math to calculate its destination, yet she can extrapolate the landing spot by watching part of its arc. The same is surely true of a cat leaping to intercept a rising bird.

I often play guitar (and, sometimes, even whistle) without any detailed idea of what I intend to do. I just watch the improvisation unfold. Sometimes, if I really like what I hear, I try to backtrack and figure out what I did. Occasionally, I can't even do that. Yet somehow what I play at one moment leads to something appropriate in the next. (Well, not every single time... )

When I say "thought", I usually mean a sort of conscious monolog, largely conducted in everyday English. That's just my habit, though. I don't think I've ever run across a strict technical definition of "thought", which is odd, considering how many books I've read on the subject.

Jerry Fodor maintains that all thought, including animal thought, takes place in a language of thought, which is an unconscious innate syntactic language, with inate symbols with built in meanings. So anything, chimps, humans, dogs, that can think and reason has a syntactic language built in as it were.

The language of thought isn't conscious, it doesn't confer linguistic abilities like ours, Fodor thinks, still something like a real full blown language underlies thought always.

Does he present any evidence for this, or is it just a gut feeling on his part? I'm trying to imagine what such evidence might look like and not having much luck.

There are some difficulties here, I believe. First, if all human beings have an innate syntactic language, one would expect all human languages to have the same grammar--the grammar of LOT. It would naturally inform our speech, and it would be positively counterproductive to speak a language with a different grammer from LOT, because it would take time and energy to translate back and forth from LOT, where the real thinking is going on.

I agree that this is a valid concern. However, in spite of the superficially great variety of syntaxes in the world, there seem to be some basic principles that are common to all of them.

[Disclaimer: In spite of what Jeff said earlier, I am not a linguist. I just read a lot about languages and linguistics, so I may well be one-hundred percent all wet.)

For example, there's that whole left-branching/right-branching thing. It reminds me of chirality in proteins and amino acids. All it takes is a little tweak at the very beginning, and everything falls into place. The rules for chemical bonds are the same in D-alanine and in L-alanine.

Actually, it's not quite that simple with language, because languages are not pure strains. Languages have messy histories, so you often end up with odd cases, such as Subject-Verb-Object languages, like English and Mandarin, in which, contrary to the "rule", adjectives precede nouns. Still the overall tendency of syntactic characteristics to appear in complementary groupings is striking.

Then there are the--possibly related--questions of language acquisition and the formation of pidgins and creoles. I've read references to studies showing that children learing even highly inflected languages tend to over-generalize and over-simplify, and may even tend to prefer SVO word order. The same tendencies appear in creoles--loss of inflection and dependence on word order to show syntactic relationships. I'd like to find out more about both of these.

Finally, notice how highly inflected languages, like Classical Latin or Classical Arabic, which are almost independent of word order, have spawned chains of descendants that seem to come more and more to lose case endings and to restrict word order. Or, compare Anglo-Saxon with Chaucerian English, and then with modern English. There is even some evidence that this happened with the Chinese languages, though it's harder to demonstrate without an alphabetical writing system to rely on.

Of course, some languages, like French, seem to be reversing this trend, so perhaps we shouldn't read too much into this apparent tendency. Nevertheless, it seems that the chances are good for generalizing all syntax back to some kind of set of fundamental principles. If that's true, then those principles might well reflect, if not a common human LOT, at least a common foundation reflecting brain organization. (This doesn't seem to far-fetched, given what appears to be a very specific division of labor between Wernicke's Area and Broca's Area.

Second, if non-human animals have a LOT, with a syntax, given their striking intelligence and social organization, all that would prevent them from talking to each other in sentences is the inability to vocalize sufficiently. They've got all the cognitive stuff built in. But, given the extraordinary advantages of such speech in socially organized communities, it is terribly hard to believe that evolution wouldn't have broken through in several places to full-blown language. Wolves would talk, chimps would talk, we wouldn't expect to be the only ones, certainly.

I think this is a pretty good argument. I was trying to come up with something similar one or two posts back, but couldn't get it to jell.

However, evolution being contingent (it can only work with what's already in place, and it can only take advantage of mutations that actually occur), we can imagine an improving mental capability that just never happens to have any organs available to express the degree of detail that that evolving intelligence is capable of handling.

Of course, intelligence doesn't necessarily have to communicate with others to be of great value to its owner. On the other hand, I can also imagine a back-and-forth between the evolution of intelligence and the development of language, with improvements in one area setting the stage for improvements in the other.

On the other hand, if LOT is mistaken, if animals have no inate grammar, well, how do they represent complex states of affairs, as I agree with you they do?

To me, "grammar" implies a mechanism for showing relationships among linguistic (or mental) objects. Is it really necessary to abstract the world into categories (word, parts of speech) at a conscious level in order to see relationships? Does a toad have to think, "Oh, there's a fly. Am I hungry? Yep. Now, let's see, which way is he headed? Is my tongue long enough to reach him?" I don't think so.

Even humans don't have to verbalize every little decision. In the late '50s, my father had to go out to some North Texas oil fields to measure some pipe that had been installed by Brown and Root, and I went along with him. We had this little measuring wheel--like a bicycle wheel with a handle, and he took advantage of my youth and sent me off across a field to measure one section. Then, he remembered that he needed to go somewhere else, so he told me to just leave the wheel where it was and go with him. When we came back, I started off down the pipe to pick up the wheel and continue with my measuring. About halfway there, there was a vigorous rattling in the grass to my right front, and I immediately levitated, ending up several feet to my left rear. I don't recall any cogitation (nor any conscious thought of "me" or "snake"), but I know that I picked the correct direction to move in, which implies that (1) I knew that there was something dangerous, (2) I knew where it was, and (3) I knew what to do and how to do it. I think it's hard to say what the limits of this style of information processing are.

I assume everyone is familar with the flocking/schooling algorithm used in the "Boids" simulation. This shows how what appears to be complex behavior can be simulated using a few very simple rules.

So we have arrived at the question:

How does thought work? Can something without a grammar nonetheless deploy concepts to represent complex states of affairs? How?

What? I read this far and you don't have the answer? Phooey!!!

[anniemcu]

I always thought that it was living in water that kept dolphins from evolving.

On land, just getting from point A to point B is a problem requiring some sophisticated machinery, and corresponding brain machinery. Land mammals ended up with hands, and I think hands are a prerequisite for evolving intelligence. Just because of their capacity for manipulation and general problem solving, as long as you have developed the brain for them.

Caj

Interesting, ... do you not think it takes a good deal of "sophisticated machinery and corresponding brain machinery" to navigate the depths and vast expanses of the seas? I dare say you would have some trouble doing it without a compass and more.

[Martin Milner]

I wonder what would have happened with Phonix if the request had been "go to the ball, and then before you touch the frisbee, touch the hoop".

That's exactly what I was saying. If you told Phoenix to go to the ball, then the frisbee, then the hoop, she did those three things in that order. If you said, go to the hoop, then the frisbee, then the ball, she did them in that order.

I don't think we're saying the same thing, though... In your first sentence (upper quote box), the correct touching order was ball, hoop, frisbee, which is the same order that the nouns appeared in your sentence. But in my sentence, the correct touching order and the noun order are different: touching order ball-hoop-frisbee but noun order ball-frisbee-hoop. I've added in bold font to highlight the different section.

I agree that if you changed the noun order then the animal's response will change to match, assuming it recognises those nouns as pertaining to particular objects. But if you *don't* change the noun order, and instead rely on syntax to communicate the changed intention, I wonder what would happen.

To put this in layman's terms (i.e. mine, as I'm not currently studying for a Master's Degree in Linguistics), Brad, Beth is suggesting that if you say to Pheonix:

Touch the Ball, then the Frizbee, then the Hoop

what Pheonix understands is

blahblah Ball blahblah Frizbee blahblah Hoop

and if you say

Touch the Ball, then before you touch the Frizbee, touch the Hoop

he understands

blahblah Ball blahblahblahblahblah Frizbee blahblah Hoop

and will touch them in the same order as before.

If he actually does touch Ball, then Hoop, then Frizbee in the second example, he has understood at a deeper level, and it would be very interesting to know if this is true.

Noun sounds for objects are the easiest part of a language to learn, and this would indicate intelligence on the level of a toddler. Understanding the more complex sentence would indicate a far greater degree of intelligence.

Regarding intelligent pets, I had a hamster who learnt to climb the stairs up and down, and came when called.

[Walden]

Regarding intelligent pets, I had a hamster who learnt to climb the stairs up and down, and came when called.

hee hee

Milner

Hamster

Stairs

[jim stone]

Oh, oh, Walden's syntactic engine is on the fritz!

[jim stone]

Consider, Darwin, the monkey with the two sticks that fit
together, who can't reach the banana with either stick,
then, finally, inserts one in the other and gets the banana.
The monkey realizes that the sticks fit together and that
if he fits them together he can get the banana. As he wants
the banana, he acts.

Probably this isn't like a frog automatically plotting
the trajectory of a fly, something much more intelligent
is going on, which we capture with words like 'realizes'
'understands' followed by a 'that' clause, e.g.
'that the sticks fit together.' 'That'clauses denote
propositions, complex states of affairs.
So the monkey grasps a complex state of
affairs--he understands a proposition.

Also there is practical reasoning going on.
The monkey realizes how to get the fruit by
fitting together the sticks, and so he does it.

Where there is reasoning, real reasoning, there
is thought. The monkey is able to represent
complex states of affairs, to grasp propositions and
reason with them.
That's sufficient for thought.

But it's unlikely that he says to himself: 'Hey,
those sticks fit together! If I put them together
I can reach the fruit!' Monkeys probably
aren't talking to themselves in that way.
Probably they lack the wherewithal,
a syntactic language.

So there is real non-occurent thought and reasoning.
And animals that lack syntactic ability
are able to grasp complex propositions,
which means they can represent
complex states of affairs.

How do they do it? How is it possible.

LOT (Language of thought) is an answer--they DO have syntactic
sentences operating unconsciously, which
represent propositions as our English
sentneces do. That language is 'Mentalese.'
We have it too, with a larger innate
vocabulary--possibly a somewhat different
grammatical structure.....

Fodor's theory has the advantage that it at least
provides an explanation of how cognition
of this sort happens. The brain is a digital
computer, and it operates in a hardwired
machine language. This is its attraction,
I suppose--it accounts for the
phenomena. The difficulties I raised
aren't decisive, I agree, though I'm
skeptical of LOT.

Note that Mentalese is going to have a grammar
as specific as that of English, say. It is a
full-blown syntactic language, not merely broad
tendencies that enable us to accept a wide
range of grammars.

This sort of theorizing is on the border
between philosophy and psychology--
it is a sort of preliminary science. Ultimately
the computer model is being deployed--how
shallwe approach the brain scientifically.
Well, let's look at it as a digital computer,
but these operate on a machine language,
so.... Cog Sci

The rival computer model in cog sci has been a
connectionist system, which has no
machine language, but I don't think I
can explain this idea clearly right now.
Paul Churchland can.

[dubhlinn]

hunger

motivation

dinner

sleep

Slan,
D.

[jim stone]

oh, no, it's catching!
Syntactitus!
Why, why won't they use rubbers?

[Jayhawk]

Jim - I find your LOT theory to be a good basis for the behaviors that Goodall and CO. have seen in the wild. Despite Darwins assertions that this is more of a response without thought behavior, family bonds remain throughout life, plots to overthrow the lead chimp can take weeks to ferment - all these things to me show memory, the ability to plan, and the ability to think several steps ahead.

It seems unreasonable to assume that animals that don't have the capability to speak or sign would evolve the complex mental structures needed for communicative language. What would the payoff be for the resources required for such a thing?

Is evolution static now? I'm pretty sure since we didn't see our ancestors make the leap between thought and language (can we say with 100% certainty which came first?). Since they're our closest living ancestors, why wouldn't they be moving towards speech? They live in social groups that war with other groups over resources, raise young much as we do, can grasp objects and make and use basic tools...why wouldn't they be moving towards speech, especially as natural resources dwindle and they need some competitive edge to stay alive as we humans alter their world.

Eric

[jim stone]

just for the record--the LOT theory is Jerry Fodor's
a philosopher of mind at Rutgers, formerly of MIT.
I'm skeptical

[avanutria]

Is evolution static now? I'm pretty sure since we didn't see our ancestors make the leap between thought and language (can we say with 100% certainty which came first?). Since they [chimps] are our closest living ancestors, why wouldn't they be moving towards speech? They live in social groups that war with other groups over resources, raise young much as we do, can grasp objects and make and use basic tools...why wouldn't they be moving towards speech, especially as natural resources dwindle and they need some competitive edge to stay alive as we humans alter their world.

"Since they're our closest living ancestors" = a link with speech ... this is a commonly held view. I recently read something about it, I wish I could recall what book it was in. If I come across it again I'll give the details. It *might* have been in "The Language Instinct" by Steven Pinker, but I can't promise. A good read regardless. Once exams are over, if anyone still cares, I'll look up the reference.

Anyway - the book I read talked about the fallacy of the argument "if humans and other primates came from the same ancestor, and humans have speech, shouldn't other primates have the same tendency/potential to develop speech?" The book suggested that the physiological developments necessary to ultimately control speech (lowering of the larynx, finer muscle control in the lips, etc) may have come *after* the human genetic line branches off from the rest of the primates. If that is the case then there is no reason chimps or any other primate would be able to develop language.

The argument "but they're closely related to us" is invalid because we're not talking about the *current* species, but the ancestors of both lines. The book posed a hypothetical situation - what if, through some sort of disease, all the chimpanzees had gone extinct thousands of years ago? Would we then look to another primate such as gorillas as our "closest living relative"? To take the hypothetical argument a step further - what if, millennia ago, aliens came to Earth with a hankering for fur coats. Hairless humans were spared but all furred animals were wiped out. Would we then be looking to mammals like armadillos and hedgehogs as our closest living relatives? We shouldn't be looking at the current species and comparing them, but the past species, the ones that ultimately developed into humans, chimpanzees, tamarins, bonobos, etc.