Saturday 7 July 2012

Marthe Kiley-Worthington: Comparing Elephant and Equine Mental Traits, Subjectivity and Consciousness

      Abstract: A combination of both reviewed scientific knowledge and knowledge gathered from philosophy of mind, critically assessed anecdotes & centuries of folk knowledge concerning the cognition of other mammals (Conditional Anthropomorphism) is proposed as a rational method to begin to outline these species subjectivity & consciousness. This paper briefly examines mammalian similarities and species differences in bodies & behaviour ( sensations, feelings, emotions learning, ecological and social knowledge, rationality, dreaming & imagination, awareness of self, theory of mind & comprehension of human language) , their  probable resulting mental attitudes, subjectivity and type of consciousness. Such an approach allows a greater understanding of another species consciousness, and can, perhaps, enrich our own.

    A comparative study of equine and elephant mental attributes leading to an acceptance of their subjectivity and consciousness . M.Kiley-Worthington. Journal of  Consciousness Exploration & Research  Jan 2011 vol 2 p 10-50.
    Animals in circuses and zoos: Chiron's world

Comments invited


  1. Yours truly, I don't understand the relevance of these presentation.

  2. Concepts are being exposed without any of the presented variables being operationalized so far. This basic conceptualization of M. Kylie-Worthington is elusive to me: nothing has been defined until now.

    I don't understand. What is the point of this presentation in elephants' consciousness if nothing has been defined? I'm thinking about that histogram on personality.

    1. Concepts? Operationalized? Defined?

      I think the speaker said she did not have time to discuss the slide on personality differences other than to name them, but they are further discussed in the paper.

      (Be careful not to be too scientisistic... That was what the speaker was calling into question, if one really wants to know whether and what animals feel.)

    2. If one really wants to know whether or not an animal feels, I don't think it should be determined from an outsider’s perspective. It would be as biased as attributing to a teddy bear feelings and emotions: we would be "projecting" (I don't like the word) our very own states of mind into it. I do not have a particular approach to solve the hard problem, apart from the fact that I still have, in light of all the conferences watched so far, "hope" that one day could emerge from the “doing” a better understanding of the “feeling”. I know your opinion, but I’m not here talking about the basic mechanisms but of a more nuanced idea of what feeling is (still not its mechanisms, but a real questioning of the notion of feeling).

      Maybe also that the “hard dilemma” is an irresolvable question. What I think is that we should not be trying to explain the feelings of any dog, or elephant, from an external perspective only as in this conference: it is a preferential track to be biased and anthropomorphically driven by our human biases. The same goes for our interpretation of other human beings' needs: our intentions might be wrong, good, spot on, out of the blue, etc.


      1. The problem of determining whether anyone other than oneself feels is called the "other minds" problem (not the "hard" problem).

      2. The only way to know whether anyone other than oneself feels is "from an outsider's perspective."

      3. The Turing Test (does it behave indistinguishably from a feeling entity?) -- whether sensorimotor, verbal or neural -- is the only evidence we can ever have.

    4. As I understood her talk, Kiley-Worthington didn't pretend that her arguments concerning horse and elephant consciousness were absolutely indubitable. As she said herself, she was only trying to shift the burden of proof in debates concerning animal consciousness. This is why the kind of arguments about behavior she adduced were useful.... I am not sure how we could try to explain the feelings of dogs or elephants other than from an external perspective.

    5. I think that that kind of observations, even if they have no "real scientific bases" are really important. Don't we do systematic observations in behavioral therapy? Isn't it how Darwin found his evolution theory? The observations that Dr. Kiley-Worthington shows of courses doesn't mean that animals have a consciousness but makes us feel like they do. This is just accumulation of evidence that points to consciousness. We could be wrong, but as Dr. Harnad said, we could be wrong about other humans too.

    6. Of course, our operational definitions of animal consciousness will be inherently faulty, as we will never know what it is like to be a horse, or to be an elephant. This absolutely does not dispel the need for observation based studies of the behaviour of sentient beings.
      One possible 'real scientific' use for this is in the creation of behavioural phylogenies amongst all vertebrates - we all too often assume our human 'doings' are exceptional.

    7. On anthropomorphism, I'll join Marjorie and quote De Waal:

      "Even though no anthropomorphism proponent would propose to apply such language uncritically, even the staunchest opponents of anthropomorphism do not deny its value as a heuristic tool"- Primates and philosophers, 2009, p.63

      We can use anthropomorphism to elaborate hypothesis and to qualify observations because we are part of the scientific process. Like Prof. Harnad pointed out: it is better to overrate than underesimate when we are talking about animals.

  3. I was wondering if you thought that animals had something like Axel Cleermans "Consciousness is the brain's unconscious theory about itself." Do you think that animals know that they are different from each other in the fraternity and act accordingly, have something that relats to a "theory of the self"?

    1. I was wondering the same thing as Thierry about animals in ‘hierarchical’ social structures and what it would mean when they act accordingly to their ‘status’ about their recognition of others in their species and their own role in the ‘society’. If they had ‘theory of the self’, what would that entail? Also, what would it mean if animals were/are able to distinguish themselves as different from other species? To me these would re-enforce a social structure in animals; however, does this strengthen our attribution of consciousness to animals? (considering a social structure does not seem necessary for consciousness)

      Izabo Deschênes

    2. Yes, I would have liked to have heard some of the speakers address 'theory of self' in animals too. Although I do not think it is necessary to have a 'theory of self' to 'feel', it would have been interesting to have heard about some of the studies on the topic - I am especially a fan of the mirror studies!

      'Self-Recognition in the Asian elephant and future directions for cognitive research with elephants in zoological settings' Plotnik 2010

  4. If experience is subjective and can be known only be the individual that experience it, can we really rely on our mammalian physical and behavioral similarities to infer the way animals feel or even to infer that animals are conscious? If we are able to create robot that behave exactly as we do, can we infer they have consciousness? Can we even know that animals are conscious and that their behavior is not the result of unconscious processes?


      Pauline, you are making a mistake. You are imagining that because one cannot know *for sure* that any other entity than oneself feels (the "other-minds" problem) it is therefore rational to conclude that animals do not feel until one has proof to the contrary.

      First, verbal report is not proof to the contrary in humans: it is just further behavioral evidence; same is true of neural correlates. So with neither of those can one be sure either.

      Second, one cannot be sure about many facts that one nevertheless takes to be true: that apples will continue to fall down instead of up; that the "laws" of science are true; that the external world exists.

      All those logical grounds for scepticism are there, in each case, but the rational conclusion, in each case, is not that therefore the fact is not true, but just that it is not certain.

      In the case of animal suffering, there is the added factor of Pascal's Wager -- which is that it you make the mistake of taking the logical grounds for uncertainty as grounds for provisionally assuming falsity, then the consequences (for the animal) are incomparably worse than assuming truth. (It's rather like assuming "guilty until proven innocent" rather than "innocent until proven guilty.")

      Empirical (i.e., "scientific") truth is a matter of probability, not certainty.

      And, as Turing correctly pointed out, in the case of feeling, the T3 (or T4) is the most evidence we can expect in the case of robots too.

  5. It was quite interesting to hear that the meanings of messages given to equines are context dependent. Can you please give some examples? If I got it right this should be some kind of evidence that equines have consciousness as usually consciousness is supposed to be involved with such higher cognitive functions.

  6. Thank you for your comments.
    First personality profiles
    As Stephen said are in the paper . I measured the amount of times each individual performed & received behaviours batched into functional categories: aggression, affiliative, interest, withdrawal & avoid and uncertainty. ( the last category formed from previous research on what these animals do in approach avoidance or frustrating situations, from experiments where they can see smell food but cannot get to it).
    The individuals were then ranked in each of these ( and the rankings correlated). The most clear correlations were that there were some individuals that were high ranking in all of these and others that were not. Thus there were individuals one could say were 'socially involved' and some who were not , like in humans.
    This was an effort to look more closely at social organisation and the similarities or differences to human language. The most important results were:-
    1) The idea of a dominance hierarchy ( which like consciousness is rarely defined and has become a metaphysical belief in ethology) did not fit well in these free ranging animals. There was no inverse correlation with aggression & withdrawal as there should have been.

    2) There were individuals who had different roles: some performers some receivers, some 'extrovert' some 'introvert' , some aggressive, some affiliative, some just 'uncertain' etc.

    3) We then looked at the number of times a behaviour what performed and the same behaviour was performed as a recipient response, ' do as you have been done by'. There were 11 of the 85 behaviours in elephants ( and more or less the same % in rhinos and horses) which were very significantly ( P.0.001) .....more likely to be done back. Does this mean anything about imitation or mirror neurons? I dont know but just a thought and something we are continuing to work on.

    4) because the dogmatic belief in the organisation of hierarchies in terms of competition did not fit here, perhaps it is not relevant here. These large herbivors do not have to compete for patched food distribution , or anything else in their day to day lives particularly, it is either available to all or to non, eg grass, leaves, shade, shelter, water etc... so why do we have to think about their organisation in terms of competition? There is no need for this, but if sharing of information about the environment ( by social learning for example) is one of the main reasons for living in a group then each individual needs to make sure he is accepted and can stay in the group. To this end we looked at 'sticking behaviour' (being nice to each other showing interest etc) and 'splitting behaviour' ( being nasty, or avoiding others). Over 70% of the behaviours in each species showed "sticking", and interestingly we also measured 'ignore' :do nothing when someone does something to you, and that was very common... perhaps a sort of deflationary strategy.

    All of which makes one wonder if we should not look more closely at the organisation of mammalian societies and perhaps evolutionary dogmas?????

    An example of 'context independent' communication is a neigh of a horse. He will neigh, and the same neigh ( we have measured 10 parameters of several 100 calles) in different situations. He may be hungry, lonely, welcoming, frightened. What he is communicating can only be assessed from looking at the context. ( Kiley 1974 paper).

    There are some behaviours that are situations specific ( eg tusking another if an elephant, = aggressive irritation) but most of the vocal and visual behaviours we measured ( we measured the responses to try and assess the meaning) were context independent ( tail wagging, head shaking, flap ears, swing trunk etc).

    Hope this helps, let me know if further points.

  7. Where did our distaste of anthropomorphism originate? I can only guess that our anti-antropomorphic view comes from experience where this approach has been incorrect. Perhaps the "conditional anthrophomorphic approach" will allow us to justify the use of anthropomorphic intuitions without falling into a trap. Can someone expand on what exactly this approach involves? We need to accurately define it if we are going to use it.

    1. Conditional Anthropomorphism is:-
      1) A recognising of similarities between related species ( mammals in this case) , until proved otherwise we should assume that they feel and are conscious and have other normal mammalian mental aptitudes. (ie if humans feel then other mammals must feel, if not why not?.... other mammals do not construct robots because they do not manipulate the world or have language. But they do feel because they care about things and look after their young , have likes and dislikes etc. ANTHROPOMORPHISM
      2) Recognizing species differences, that the elephant is very much larger, has a trunk, the horse runs faster & has 4 legs and eats grass... all the differences in body , sensory systems, physiology, anatomy, behaviours etc etc that we know about. CONDITIONAL PART
      3) The 3rd part which not got to yet is the individual's different world view as a result of his passed experiences and genetic component, he may or may not have the same view of the world as many of the rest of his species. For example if a horse has only 3 legs he will have a different concept of space and speed from his con specifics.

      I did say all this in the talk by the way, sorry it was not clear?

    2. Thank you for taking the time to explain again. The talk was clear, I just didn't catch everything the first time (too many talks in one day!) Its great that we have the recordings to look back on.

  8. I greatly appreciate the topics which were raised in this talk and some of the profound parallells which were drawn.
    I think it is important, however, to pay attention to what we choose to assign as signs of conscious behaviour - simply because WE regard it to be a conscious trait in ourselves does not necessarily mean that it is a mark of consciousness.
    The example I was thinking of specifically was MKWs suggestion that the presence of complex social societies was an indication that theory of mind might be at work; bees and ants for some of the most highly ordered, complex, and hierarchical animal societies.
    Yes, I understand that her argument revolved around 'lack of evidence for is not evidence against', but if we are going to look two rungs down the evolutionary ladder at behaviour, we've got to go all the way down...

    1. But bees are not mammals are they? Invertebrates may have TOM and other things of that nature but they will be doing it very differently. Do you really think that a weasel and a bee have a similar consciousness/feelings, and that this is quite different from yours? We know something about weasels, let us start there.

      Bees read particular ritualized messages from others, but I am not convinced they know that the other is a mind and has feelings and desires. I just think we need to start somewhere rather than switching off from others. .. it may be that the world has a collective consciousness who knows... we cant rule that out.

      Equines/elephants/weasels social life is very complex and in many ways similar to ours, not like a bees, and it is very difficult to work out a system how this would work if they were not conscious in the sense of having some idea of another feelings.. they also make mistakes do bees?

  9. Xavier Dery ‏@XavierDery

    I like the distinction M. K-W makes between folk belief and folk knowledge, the latter lenghly confirmed and in line with evolution #TuringC

    2:20 PM - 7 Jul 12 via Twicca Twitter app

  10. Looking at her paper, I really liked Dr Kiley Worthington's approach to animal personalities. This is the start of a really interesting field and opens the possibility to draw parallels between species. I might be afraid it is impossible to get out of our human perspective while exploring other animal's subjectivity though...