Resurrecting the thread for another anomalous case that I would like you guys to look at.

This time, it's about the alleged language capabilities of dolphins as studied by marine biologist Louis Herman. Here is the relevant excerpt from Wikipedia which explains the discovered capacities of the dolphins:

"His 1984 paper on animal language (Herman, Richards, and Wolz, 1984) was published in the human psychology journal Cognition, during the anti-animal language backlash generated by the skeptical critique of primate animal language programs by Herbert Terrace in 1979. The key difference with previous primate work was that the dolphin work focused on language comprehension only. The problem with researching language production was the issue of scientific parsimony: it is essentially impossible to verify that an animal truly understands its own artificial language production. This problem is eliminated with language comprehension studies, because the researchers control the form of the artificial language, and need only observe the behavior of the animal in response to the symbol sequence. Other controls included the use of a blinded observer who was not aware of the sentence given to the dolphin, as well as the balanced presentation of possible word/symbol combinations. Most importantly, the dolphins were tested on their responses to novel sentences they had never before been given, to test for concept generalization. Also, the dolphins were tested in novel sentence grammars and anomalous grammars as well, demonstrating that the dolphins' comprehension was not limited to a finite-state (slot-based) syntax."

In addition, the dolphins showed evidence of comprehending the grammatical structure of  subject-object-direct object  when being given commands.

So what do you think we should make of it? Do these above mentioned linguistic comprehension tests show how dolphins may be in possession of a rational intellect? Or is there somewhere an equivocation between human intelligence and dolphin intelligence, such that the dolphins need not have been universalising concepts but could have only been generalising particular images and associations?

Feedback would be very appreciated!

Last edited by aftermathemat (4/23/2018 9:46 am)

Greg wrote:

People who have young children will sometimes say things like, "She knows over a dozen words now." But these ascriptions are somewhat generous. To take an example from Stanley Cavell, we will say that a child knows the word "kitty" when she identifies her family's kitty, when she identifies kitties in bedtime stories, etc. But then she might go outside and call the neighbor's dog a kitty. Does she know what a kitty is? The bar for concept possession is, shall we say, considerably lower for toddlers and for parrots than it is for adults.

I agree with this. I know of a following case: There's a toddler in the family. Uncle Eric comes to visit occasionally. The child hears how people greet uncle Eric and the child is told to greet uncle Eric also. Then one day a whole different visitor arrives. The child greets the visitor as "uncle Eric". Evidently, the child has only learned an automated response, not the proper concepts "uncle" or "Eric".

In 30 years of training, a parrot can learn to parrot pretty impressively, but it's still just parroting.

aftermathemat wrote:

So what do you think we should make of it? Do these above mentioned linguistic comprehension tests show how dolphins may be in possession of a rational intellect? Or is there somewhere an equivocation between human intelligence and dolphin intelligence, such that the dolphins need not have been universalising concepts but could have only been generalising particular images and associations?

Feedback would be very appreciated!

A sympathetic summary under the spell of amazement is not enough. The full story (the original research paper in its entirety) is needed.

Last edited by seigneur (4/24/2018 3:24 am)

About this:

Callum wrote:

he briefly described just how much Albert the great and Aquinas gave to animals and That, though the terminology will be different, current headlines on dolphins, elephants, higher apes etc can all be described in the sense Aquinas did.

And:

seigneur wrote:

A sympathetic summary under the spell of amazement is not enough. The full story (the original research paper in its entirety) is needed.

 Animal Minds: Beyond Cognition to Consciousness  (https://books.google.com/books?id=K2uXAwAAQBAJ&printsec=frontcover)  is a book that describes some of the results in more detail.

Here are the relevant excerpts from the 2013 expanded edition, which can be searched manually by typing in the keyword dolphin. Below is a sample of Chapter Twelve on Dolphins and Apes, page 229:

""But recent experiments by Mercado et al. (1998, 1999) demonstrate that bottle-nosed dolphins understand and comply with the command to repeat one of their previous actions when that action is described by human gestures that designate the action to be repeated in terms of objects or locations, including the command  "Repeat what you just did"  or  "Do something new and different." ""

Here is something relevant from page 232:

""Herman (1986, 1987) and his colleagues have approached the question of dolphin cognition by means of an intensive training program designed to assess the degree to which these animals can understand not only individual signals but simple combinations of signals that are related to one another in a manner resembling the grammatical rules of English. This emphasis stems from the widespread conviction that combinatorial productivity based on the use of rule-governed combinations of words constitutes an essential and unique feature of human language. Herman's experiments have concentrated on comprehension or receptive competence of dolphins rather than on their ability to produce communicative signals. He trained one female bottle-nosed dolphin, Akeakamai (or Ake), to respond to gestures from a trainer at the edge of a tank, and another, named Phoenix, was trained to respond to underwater whistle-like sounds. Both dolphins learned to respond appropriately to about thirty-five signals, which included the names of the objects in the tank, and actions such as toss, swim under, jump over, fetch, and put something in or on top of something else. They also learned a few modifiers such as right and left, surface and bottom. A typical combination of signals was  RIGHT PIPE TAIL-TOUCH,  meaning that the dolphin should touch the floating pipe to her right with her tail.""

""Both Ake and Phoenix learned to respond appropriately to numerous combinations of these commands, which could be as complex as five-unit sequences such as  PLACE BOTTOM PIPE IN SURFACE HOOP.  In the system used in these experiments  PLACE IN  is a single command, and the order of the signals differs from customary English but nevertheless constitutes a set of rules in which the sequential arrangement adds important meaning to the individual signals themselves. This much was not especially novel, because many dolphins have been trained to carry out sequences of behaviour of greater complexity. But in their training, these two dolphins were never presented with more than a small proportion of the meaningful combinations of these commands; the others were reserved for tests of their comprehension of these rules that governed the combinations. These rules were comparable to those of English, in that the order of the words or commands determined which would be the direct or indirect object and which modifier applied to a particular command. """

Here are some very important comments on assumptions about human language and sign combinatorials that are adapted to animals, taken from page 234:

""The great emphasis on syntactical rules as a fundamental property of language has been developed primarily by scientists and scholars whose native language is English, one of the few human languages in which word order provides almost all of the syntax. But in most languages inflection of words is used to convey the grammatical relationships between them, and English does retain a few vestiges of inflection, for example, the possessive form of nouns and the different forms of pronouns such as he, his, and him. The fact that so many human languages rely heavily or primarily on inflection of words to convey syntax suggests that inflection may be a more basic and, perhaps, a more easily utilised way to express grammatical relationships. If so, attempts to teach a combinatorially productive language to animals might be more successful if they employed inflection rather than word order.""

Which prompts me to ask: If animals do indeed achieve a better level of combinatorial production when being taught language because they employ inflection, would this actually be a sign of rational intellect, or can this too (if such attempts of teaching animals inflection in combinations were to succeed, which is likely) be interpreted as using simple imagination, association and generalisation, rather than true universalisation and intellection?

---

Then we have another excerpt which contrasts the capacities of the dolphins with those of humans, which is also highly relevant. From the same page as the above:

""A more fundamental distinction between human language and what dolphins and sea lions have accomplished, at least so far, is the ability to switch back and forth between comprehension and production of communicative signals. This is obviously a key attribute of human language, and it is also an ability that has been attained by some Great Apes...But we do not know whether dolphins could learn to comprehend and produce signals interchangeably, and therefore this question remains open. Obviously dolphins and sea lions cannot be expected to produce human gestures, but Phoenix, who learned to comprehend the rules relating commands presented to her as underwater sounds, could almost certainly learn to imitate those sounds. Appropriate experiments with other dolphins or sea lions trained to communicate by sounds might reveal whether they are capable of using acoustic sounds interchangeably in both the receptive and the productive modes.""

If we combine the above excerpts with the results as summarised by Wikipedia of their ability to recognise and generalise  "concepts"  when understanding and responding to novel sentences, as well as comprehension in novel and anomalous grammars which shows them not being limited by finite-slot syntax, it seems we have quite the intelligent animal on our hands! 

But whether or not that means they have a rational intellect is the key problem here.

What do you think?

Last edited by aftermathemat (4/24/2018 7:43 am)

seigneur wrote:

Sorry, but all this underlining, bolding, and capitalizing makes me highly suspicious. Are those the keywords you searched?

I personally never go by keywords. I go by definitions and step-by-step outlines of procedures. I go by taking first things first, like, How do these guys define "understanding"? How do they ensure they experimentally isolate exactly what they defined? Etc.

These weren't exactly the keywords I searched. I typed in  "dolphin"  so as to isolate the part of the book that talks about dolphins, and then went on to quote the most interesting excerpts. 

The underlining and boldening were all done by me to highlight the most important points of interest.

As for how they define understanding, I think one of the excerpts above focused on syntactical manipulation as the definition of what is essential to human language, and then searched for such an ability dolphins, and this evidence then shows how dolphins can comprehend the grammatical / syntactical structure of what they are hearing.

seigneur wrote:

From what I can see from the text, the authors have a very nebulous idea of language. In reality, syntax only becomes a thing given a fairly sizable vocabulary, say a 100+ words that represent all word classes. It's absolutely key that the words should represent all word classes. From what I see in the text, they have 35 signals and no explanation of their grammatical status.. There is some outright misuse of linguistic terms like "modifiers", "action" (apparently meant to be "verb"), "direct or indirect object" - i.e. lots of willingness to see linguistic behaviour in those animals, but no appropriate rigour to ensure it.

Everything that the dolphins actually do there seems to fit just fine under basic circus trick logic: When I do A, you do a. When I do B, you do b. That's it.

So what I gather from this is that the animals weren't in fact taught any meaningful syntax since true syntax requires more than a 100 words that include all word classes, whereas the syntax the dolphins were taught was rather meager and poor. 

As for the abuse of linguistic terms like  "modifier",  "action"  and  "direct object", when I first read the above excerpts my first thought was that this wasn't really that impressive. The dolphins used simple memory and imaginative association to recognise the intention that the various whistling sounds had, and then simply combined the various associations they knew the whistling sounds had into one coherent whole in their head. 

Now the most impressive thing about this is the claim that they knew the order in which the sounds are to be assembled and could even generalise their associations when faced with novel commands and grammars, thus showing how they can comprehend the artificial language.

But even if they recognised the correct order in which the commands should be given and even expressed confusion when the commands were given out of order, and even had something analogous to comprehension, this still seems like something that could easily be explained by their imaginative association. As for their ability to generalise their understanding of syntactical connections beyond finite-slot syntax, this still seems to me like something they could have easily done using their imagination and simple algorithms, especially given how cetaceans like dolphins have the biggest brain-to-body ratio other than humans, and how their brains are fairly developed and are naturally very intelligent, with only some parts of their brain lacking the organisational structure that primate and human brains have.

What do you think?

seigneur wrote:

This seems to presuppose that languages without inflection (Chinese, Vietnamese) require or stimulate less intellect than languages with inflection. Wrong presupposition. For example, they over there had writing long before white men did. And history of modern English is marked by drastic reduction of inflections (compare with German and Icelandic).

Otherwise seems like a fair if-then. If animals did human language, they would probably deserve human rights. But there is no sign they'd do human language any time soon.

I'm pretty sure that the trainers over at Sea World would really hesitate to give dolphins rights, or to ensure that they have at least some primitive court system to ensure that all fellow dolphin-murderers get capital punishment.

This perhaps shows how we have innate knowledge that other animals just aren't rational animals, even though materialists would dismiss the intuition by simply saying that we as more developed animals don't care about those who are lesser than us just like fish don't care about the shrimp they are eating, and more or less for some made-from-whole-cloth evolutionary reason.

But anyways, MacIntyre once wrote of how we should define rationality not strictly by use and capacity of true language, but by using the language to self-reflect on one's decisions. This would be what is called the  "age-of-reason",  which I'm pretty sure dolphins don't have, even if we concede that they had language because they can comprehend all syntax. 

And let's not forget Chomsky's valuable insight of how comprehension of language should eventually lead to use of language amongst animals that have it. Dolphins, like apes, simply don't use syntax and the rules of English in nature and clearly don't have true language out in the wild, so to find out they can in fact master language would be like finding a flightless bird that had functional wings. It's simply implausible, and suggests that, whatever power they are using to manipulate syntax, it ain't intellect.  

This is also relevant because the excerpts mention how Great Apes have the ability to switch between comprehension and production of communicative signals, which dolphins and sea lions don't have (or at least not yet, since they are hoping they will turn out smarter in future experiments and basically have conversations with you). And since Great Ape language has already been criticised by linguists such as Chomsky and Pinker, the fact that dolphins are below apes in terms of understanding should give us a breather here. And the fact that sea lions are also being considered as having language along with dolphins should give us pause. They're clearly not rational animals, and to suggest otherwise invites speculation of how many other animals could potentially  "learn language"  as well, which undermines the point of attempting to prove other animals as being like humans in possessing language.

Last edited by aftermathemat (4/24/2018 12:45 pm)