On Bateson's Problems in Cetacean and Other Mammalian Communication

Bateson, Gregory 2000 [1966]. Problems in Cetacean and Other Mammalian Communication. In: Steps to an Ecology of Mind. Chicago; London: University of Chicago Press, 364-378.

Introduction
This text began as mere notes on Bateson's paper. I re-read his "Problems..." with a measureable will to understand it better and the text quickly turned first into a review and then a reflection on it until it became just a playground of to fiddle with different aspects of Bateson's work. I think the main motivation behind it was just me finally, after several years of admiring Bateson at a distance, getting into the thick of his approach to communication. I would say it is one of my favourite standpoints on communication and very much compatible with other approaches in semiotics. I do think that the problems that Bateson discusses in this paper are some of the most interesting ones put down on paper and will mostly continue to be interesting when read from a computer screen.
The tone of this text is informal and because it is intended to be read from a computer, the references are hyperlinks. I took the liberty (a kind of semiotic freedom) to do write unconventionally in academic context and on academic topics. Some of the sub-headings that follow Bateson's paper on cetacean and mammalian communication and quote the relevant parts of in, but there are some detailed sub-headings in which I just have fun with it.

Transactional analysis of behavior
According to Wikipedia, transactional analysis (TA) is "an integrative approach to the theory of psychology and psychotherapy". It supposedly integrates elements of psychoanalytic, humanist and cognitive approaches and was first developed by Eric Berne. TA is, among other things, "a theory of communication that can be extended to the analysis of systems and organizations". Although Bateson was not himself a psychiatrist, he did work in the 1950s with Jurgen Ruesch, a semiotically inclined psychiatrist (or as he described his own approach, a "social psychiatrist").

Deutero-learning
Deutore-learning is defined briefly as "learning to learn". This concept was influential in the work of the so-called "Palo Alto school", a group of researchers engaged in studying schizophrenia. Their brand of "communication psychiatry" was also linked with the work of Ray Birdwhistell and Kenneth Pike. In the paper under discussion, the transactional analysis of behavior is briefly summarized in terms of deutore-learning:

1. "a relationship between two (or more) organisms is, in fact, a sequence of S-R sequences (i.e., of contexts in which proto-learning occurs)";
2. "deutero-leaning (i.e., learning to learn) is, in fact, the acquiring of information about the contingency of patterns of the contexts in which proto-learning occurs"; and
3. "the "character" of the organism is the aggregate of its deutero-learning and therefore reflects the contextual patterns of past proto-learning."

This technical summary can be "intralingually translated" if we replace some "outdated" terms with closer-to-home semiotic jargon. For example, "a sequence of S-R sequences" could very well be replaced simply with "communication", seeing as communication in fact does consist of a "transactional" relationship between two (or more) organisms that engage in an exchange of signs. Moreover, the "S(timulus)-R(esponse) sequences" accord to "functional circles" (Funktionkreis) in Uexküll's theory. That is, what we have at hand is in fact a sequence of Receptor-Effector cue circulation. Thus:

1. Learning to learn occurs in the context of communication, a relationship between two (or more) organisms that engage in an exchange of perceptual and operational cues.

That organisms learn to learn in communication with other organisms seems self-evident. That is, organisms can learn on their own, but in order to learn how to learn they must come into contact with other organisms and perhaps observe how learning is performed. Albert Scheflen, another semiotically inclined psychiatrist, put it this way: "Some kinds of teaching are carried out in twosome and rely primarily on demonstration." (Scheflen 1972: 22) In his example, this involves physical tasks (what he calles actonic behaviours), such as a father teaching his child how to hammer a nail - effectively teaching him how to learn to nail.
Understanding the "contingency patterns of the contexts in which proto-learning occurs" is more difficult. Elsewhere in the Steps to an Ecology of Mind Bateson avails that "the contingency pattern which describes the relation between "stimulus" (CS), animal's action (CR), and reinforcement (UCS) is profoundly different from the contingency pattern characteristic of instrumental contexts of learning." (Bateson 1972: 293) At the moment I cannot get a hold of Bateson's paper on this issue, titled "Social planning and the concept of deutero-learning", and written as a response to Margaret Mead's (his wife's) paper "The Comparative Study of Culture and the Purposive Cultivation of Democratic Values" (as noted on Wikipedia). In any case:

2. Learning to learn consists of aquiring information about the relationships between perceptual and operational cues in the relevant context.

In the third point summarising deutero-learning, Bateson makes a grand statement that the "character" of the organism is the aggregate of its deutero-learning and therefore reflects the contextual patterns of past proto-learning. At this point it is relevant to induce the difference between proto- and deutero-learning. According to SpringerReference, these are simultaneous levels of learning. The example positions learning how to ride a bike, learning a language or repairing a car as proto-learning; and at the same time "you are also learning something about the world and something about how things occur". That is: "deutero-learning describes the context in which (proto-)learning processes occur." For example, while learning to ride a bike you also learn something about balance, velocity, and other physical aspects of the world in which you ride your bike. Succinctly, this induces the actor's relationship to its world - that learning is not somehow separate from the world you learn in, but is about the world you learn to do something in. Thus, returning to Bateson's third summarising point, the "character" of the organism is "an aggregate of its deutero-learning" is the vague sense of what it has learned about the world while learning to do things in this world. The second part is especially releavilng: the "character" of the organism reflects the "contextual patterns" of past proto-learning; it reflects not only what it has learned, but how it has learned (in what contexts). Thus:

3. The character of the organism is it's relationship towards its Umwelt and it reflects the contexts of communication and in the contexts in which it has learned something.

I'm sure countless practical examples of animals not only learning to do something but the habits formed through learning reflecting something about the context of it's learning - e.g. where it has learned to do something, with whom (conspecific or interspecific contexts), and how etc.

Digital and analogic
This distinction has a significant bearing on semiotics. It's uncanny resemblance to Lotman's distinction between discrete and continuous sign systems has been noted by many (and Montana Salvoni recently wrote a masters thesis about it in our department). Although Lotman probably derived it from Jan Baudouin de Courtenay, a Polish linguist, who among other things coined the term morpheme and worked in Dorpat (as Tartu was then called) between 1883-1893, I also have my own theory of how Susanne Langer's (1942) distinction between discursive and non-discursive symbols influenced Lotman. In any case, the digital/analogic distinction is quite interesting and once again, Bateson's colleague Jurgen Ruesch was behind this distinction - he derived these forms of codification from Langer's distinction (Ruesch & Kees 1956: 8-9).
Bateson notes that the three summarising points on deutero-learning above are "essentially a hierarchical structuring of learning theory along lines related to Russell's Theory of Logical Types", but continues to note that these are "primarily appropriate for the analysis of digital communication". One could only presume that this is because Russell's Type Theory is about mathematics, logic and computer science and thus inherently digital; that is, in Langer's sense, related to propositions. Analogic communication on the other hand is non-propositional. Interestingly, Bateson also puts forth that there may be "systems that combine the digital with the analogic" - what Juri Lotman would call, after Roman Jakobson, syncretic verbal-visual models (Lotman 1988: 127; in Pärli 2000: 210), e.g. a "combination of several [different] sign systems" (Torop 2005: 166).
Bateson surmises that studying the communication system of dolphins is worthwhile because it is "almost certainly of a totally unfamiliar kind" and can thus "close gaps in our theoretical knowledge of communication". This is indeed brilliant, because the study of human nonverbal communication (by Ray Birdwhistell, for example), went somewhat astray exactly because it is not different enough from natural/verbal language; and in part because of this spawned "a conception of unmatched vulgarity": body language (Sebeok 1975: 12).

Communication
Bateson defines communication as the animal's behaviour "looked at as an aggregate of data perceptible and meaningful to other members of the same species". That is, communication is defined as perceptual cues that have a species-specific "tone", in Uexküll's parlance. E.g. "the inner world of a dog has already composed the specific tones (Töne) that trigger its multiple possible actions" (Chien 2006: 68). The sense, meaning or tone that an aggregate of perceptible data carry relates to a certain performance - in crude terms, an action (by a conspecific) is followed by a "meaningful" reaction.
He then puts forth a definition of meaningfulness. The first aspect of this definition is quite simple: the behaviour of one animal "affects a recipient animal's behavior". The second aspect is infinitely more difficult to grasp: "perceptible failure to achieve appropriate meaning in the first sense will affect the behavior of both animals." That is, when communication is effective then one animal influences the behaviour of another animal; when communication is ineffective then the behaviour of both animals will be affected. He stresses this point because in interspecies communication (especially between man and dolphin), "a very large portion of the behavior of both organisms is determined by this kind of ineffectiveness."
In order to comprehend the import of this aspect it seems pertinent to try to translate it into more familiar terms. The easiest way to go about this is to use his human analogy: "What I say to you may be totally ineffective, but my ineffectiveness, if perceptible, will affect both you and me." That is, when communication is effective, the signs exchanged in the process will affect our behaviours in predictable and understandable way. But when communication is ineffective, the signs we exchange will lead to unpredictable and mis-understandable behaviour on both parts - we will be mutually confused. For this to happen, though, the ineffectiveness has to be perceptible. This latter point is relevant for it qualifies the communication as true communication (as opposed to unilateral communication) - both parties must be responsive and perceptive of each other's signs.
The difference between inter- and intraspecific communication is important in this regard, as humans, for example, have verbal language and can point out the ineffectiveness by applying Jakobson's metalingual function, for example, inquiring "what did you mean by that?" etc. Ineffectiveness determines interspecific communication between man and dolphin because there is no common code - the dolphin's communication system "is almost certainly of a totally unfamiliar kind".

Pressing down the head
In an effort to be as thorough as possible, I will also review Bateson's examples of animal communication. The first illustration touches upon one animal pressing down the head of another (with open mouth on the back of the neck). There are several iterations of this "gesture" in the canines. When a coyote, dingoe or domestic dog puppy asks for milk, the mother presses the puppy's head down to the ground until the puppy stops asking. In wolves the puppies are weaned off milk and graduated to regurgitated food. The method for doing so is similar - the puppy's head is crushed down. In wolves it is not only the mother who performs this gesture, but other adults of both sexes do it as well.
The same gesture is illustrated in adult wolves in the Chicago pack. The pack leader prevents other males from "getting" the females in heat. Bateson reports that in the preceding year one other male succeeded in establishing coitus with a female, but got locked in the female. The pack leader rushed up and instead of tearing the locked-in male to pieces as we would assume from a human, he "pressed down the head of the offending male four times" and then simply walked away.
Bateson explicates the implications for research from this illustration. In his opinion the sequence is not describable (or only insufficiently describable) in S-R terms. The pack leader "does not 'negatively reinforce' the other male's sexual activity" but "asserts or affirms the nature of the relationship between himself and the other". Pressing down the head is not a equivalent to a statement such as "Don't do that." but would rather be put into human words as "I am your senior adult male, you puppy!"

Communication about relationship
Another, briefer, illustration involves the feline "asking" for food or milk. In doing so she makes "movements and sounds that are characteristically those that a kitten makes to a mother cat". Put into words the message would not read "Milk!" but something like "Mama!" (or even "Dependency! Dependency!").
These illustrations should bring home the generality of this view: the mammal "talks in terms of patterns and contingencies of relationship" and "asserts or affirms the nature of the relationship between himself and the other". Bateson goes on to discuss how the innovation of human language was not "the discovery of abstraction or generalization" but of "how to be specific about something other than relationship". The underlying assumption is that pre-verbal humans graduated from the mammalian communication about relationship to communication about other things. In Roman Jakobson's terms, we evolved from emotive, conative and phatic functions of communication to referential, poetic and metalingual functions.
Although we have acquired fancy "linguistic tricks", Bateson affirms that our "mammalian ancestry is very near the surface", meaning communication about relationship is not lost in humans, but very much present in our everyday communication. The illustration goes as follows: A says to B, "The plane is scheduled to leave at 6.30" but instead of accepting this remark as merely about the time of departure, B "devotes a few neurons" to the question, "What does A's telling me this indicate for my relationship to A?"

Metacommunication
Because communication about relationship is so close to the notion of metacommunication and Bateson worked together with Ruesch at Palo Alto, it seems appropriate to review the notion of metacommunication. The primary source that should be consulted is Communication, the social matrix of psychiatry (1951) co-authored by Ruesch and Bateson.
The first appearance of metacommunication in Communication... is political: the leader of an American party has to maintain integration in a group with diverse opinions about matters about which he must make decisions. Metacommunication is here framed in terms of nonverbal feedback to a verbal utterance: "Whenever he speaks, each utterance is a trial balloon, and he continually watches those behind him to see how far he can go". They affirm that in conventional (psychiatric) terms this is called "reality testing" but propound a viewpoint that this is more like "asking an implicit question about his own statements" or asking "What effect will my utterance have upon relations between my supporters and myself?" (Ruesch & Bateson 1951: 152).
Although metacommunication is defined as "a communication about communication", the actual core of the matter seems to be about relationships. In this illustration the party leader is ipso facto making statements like "I lack certain information about my relationship with my followers" and "I feel a need for this information" (ibid, 152). To re-use the human illustration above, this would amount to "What does the nonverbal feedback to my utterances tell me about my relationship to my voters?"
The next time metacommunication appears in Communication..., it has to do with animal communication - because of which I shall dwell on for a while. They arrive at animal communication by discussing a characteristic of interpersonal communication systems that has heretofore been neglected: "the real existence of the group as a determinant of the actions and communications of the separate persons"; that is, the fact that the real existence of a group is determined by whether each participant is aware of the perceptions of the other - or in human terms, "If I know that the other person perceives me and he knows that I perceive him, this mutual awareness becomes a part determinant of all our action and interaction", and "The moment such awareness is established, he and I constitute a determinative group, and the characteristics of ongoing process in this larger entity control both individuals in some degree" (Ruesch & Bateson 1951: 208).
Next they infer that there is little information about the evolution of the "group" in this sense, but suggest that the question of such evolutionary history is worth considering, even if only to affirm that groups that are defined "in terms of mutual awareness of perception" are something different from groups "determined merely by mutual irritability or responsiveness" (ibid, 208). They refer to Tinbergen's case of the sticklebacks as a group determined by irritability and responsiveness, but show no evidence which would indicate that either individual is aware of the other's perception. They go on to note that von Frisch's case of bee communication similarly demonstrates no reason to believe that such awareness occurs and put forth that this evolutionary step probably occurs only among primates and domesticated animals - and, that the matter needs critical investigation.
The really interesting part is their "operational" definition of a group of this higher order and propose some aspects that should be observed. The primary aspect that deserves attention seems to be what Albert Scheflen (1972: 105) calls self-censure. Here it manifests itself simply as "whether each participant modifies his emission of signals in a self-corrective manner according to his knowledge of whether the signals are likely to be audible, visible, or intelligible to the other participants" (Ruesch & Bateson 1951: 208-209).
This kind of self-correction is supposedly unusual among animals (and they do not neglect to add that in men it is desirable but not always present). They suggest identifying the following types of signals among animals:

1. "signals whose only meaning would be the acknowledgment of a signal emitted by another";
2. "signals asking for a signal to be repeated";
3. "signals indicating failure to receive a signal"; and
4. "signals which punctuate the stream of signals".

The list could go on but, these signals should prove or disprove whether the animals have a group of a higher order on the basis that "with complete awareness of the other's perception an individual should stop repeating a signal after it has been received and acknowledged by the other individual" (Ruesch & Bateson 1951: 209).
Although this may seem disconnected from our current topic, they verify that "these criteria for the existence of mutual awareness build together to give a picture of the entirely new order of communication which emerges with this awareness" (ibid, 209). This new order of communication is here introduced as metacommunication. This "communication about communication" involves all exchanged cues and propositions about (a) codification and (b) relationship between the communicators. Moreover, they assume that majority of propositions about codification are also implicitly or explicitly propositions about relationship and vice versa, so that no sharp line can be drawn between these two types of metacommunication. Thus, metacommunication is both communication about (the codification of) communication as well as communication about relationship. We can now return to Bateson's 1966 paper on cetacean communication.

Complexity of communication
Bateson's earlier dabbling with the "higher order" of (animal) groups is a perfect segue to the topic of complexity in communication. The paragraph we left off to discuss metacommunication begins with Bateson's surmising that he expected dolphin communication to manifest "the general mammalian characteristic of being primarily about relationship". By now we have a somewhat clearer picture of the evolutionary history of communication about relationship: sticklebacks and bees presumably lack communication about relationship, in mammals and especially primates this is the most prevalent function of communication and in humans it is just beneath the surface of linguistic communication.
As he so cleverly put it, one can suppose that large-brained creatures were, at some evolutionary stage, "unwise enough to get into the game of relationship". From this he deduces that there must have been some survival value for a species in getting caught in this "game of interpreting its member's behavior toward one another as relevant". He then reasonably expects to find "a high complexity of communication about relationship among the Cetacea" - they are, after all, social and large-brained.
In the next section, titled "Methodological Considerations", he moles over the difficulties introduced by this hypothesis. The problem is that of testing animal "psychology". The latter includes such aspects as intelligence, ingenuity, discrimination, etc. He lists four steps of a simple discrimination expriment:

1. "The dolphin may or may not perceive a difference between the stimulus objects, X and Y."
2. "The dolphin may or may not perceive that this difference is a cue to behavior."
3. "The dolphin may or may not perceive that the behavior in question has a good or bad effect upon reinforcement, that is, that doing "right" is conditionally followed by first."
4. "The dolphin may or may not choose to do "right", even after he knows which is right."

He goes on to discuss some "methodological considerations" about dolphins reaching one or other of these four steps. I think we may courteously dismiss this discussion in favor of something more creative - comparing this list of steps with types of signals indicating self-correction in Communication... (1951). The similarities may be coincidental but my hope is that they will reveal something deeper about Bateson's "methodological consideration" (viewed diachronically).
Whereas in Communication... the first type of signals to look out for in relation with "higher order" awareness (henceforth referred to as "(a)" for sake of brevity) is actually quite comparable to the first step in the current list (henceforth "(1)"). While (a) concerns the the presence of signals that acknowledge a signal emitted by another, (1) involves the dolphin's ability to perceive a difference between two stimulus objects. The commonality consists in discriminating a signal (a) and a stimulus (1).
Moving on to the next pair of items we discover that (a) actually accords more to (2), but then we'd have to elaborate (a) as not only acknowledging a signal emitted by another, but also perceiving it as a cue to behaviour. The commonality is here captured in the notion of communication - not only discriminating a signal or a stimulus, but also responding to it.
Given that the pairs of alphabetical and numerical lists are at this point off by one, we may next choose to compare (b) and (3). Type (b) is a signal that asks for a signal to be repeated. Only tangentially but still with a distinct possibility this "Say again?" (b) can be compared to (3) - the dolphin may or may not perceive that the behavior in question will lead to fish.
Upon review it appears that (c) also accords to (3). In this collision of items the dolphin does not perceive that the behavior in question has a good or bad effect (3) and may give a signal to indicate a failure to receive a signal (3). E.g. "Do you know what you have to do?" asks the trainer; "Nukhuh" responds the dolphin.
And finally (d) and (4) match up perfectly: the dolphin may choose to refuse to cooperate (4) and gives a signal that punctuates the stream of signals (d). Although these lists do not fully match up, they do share some commonalities and avail some general points of interest in studying animal communication: whether the organism discriminates communicative signals (a, 1, 2); whether the organism understands the signification of the given communicative signals (3); whether the organism can give feedback about it understanding or not understanding the significance of the communicative signals (b, c); and whether the organism complies with the instruction communicated (4) or can communicate the punctuation (ending, finishing) of the interaction (d). I think I am now ready to review Bateson's own discussion of the steps of discrimination.
The first thing he points out is that if the dolphin succeeds in the first three steps then it arrives at "a further choice point" with "extra degree of freedom" - it can choose not to cooperate. For Bateson, this is a point that should be investigated first, and for methodological reasons. Namely, it is important to know whether the organism is capable of step 4 because it basically decides the nature of the experiment. For sake of analogy he points out that in experiments with humans no special care is taken of step 4 because if the human subject is deemed "cooperative and sane" he usually responds "by repressing most of his impulses to modify his behavior according to his personal view of his relationship to the experimenter".
It is interesting how Bateson manages to wiggle communication about relationship into this account of methodology. The point is that being cooperative and sane implies fairly constant relational rules and the experimenter does not need to worry about changes in those rules. That is, an uncooperative and/or insane subject could very well change the relational rules - decide to work against the experiment, for example. He is obviously aware of this very possibility because at Palo Alto he did have to work with exactly these kinds of subjects - which is probably why he is not reluctant to compare a dolphin to a noncooperative, psychopathic or schizophrenic subject (I especially like the option of a naughty child). Bateson finds the ability of the animal to operate at this relatively high level of communicational complexity most fascinating, and rightfully so, for it ventures into one of the most interesting areas in animal psychology, in my opinion (succinctly put, it is the question of semiotic freedom).

Animal training
From conversations with highly skilled animal trainers (of both dolphins and guide dogs) Bateson formed an impression that "the first requirement of a trainer is that he must be able to prevent the animal from exerting choice at the level of step 4". To put it more crudly: that the trainer must be able to coerce or otherwise influence the animal to cooperate so that there would be "no nonsense about it". Read from a more playful position, it is heartening to point out that animals do make "nonsense" about being subjected to human expectations. Especially in the inherently semiotic sense of "nonsense", meaning non-significant behaviour, it almost seems to indicate that animals can not only choose not to "do 'right'" but choose to do something completely incomprehensible to humans. I find there is something poetic about this idea.
"In other words," he goes on, "it is a primary condition of circus success that the animal shall abrogate the use of certain higher levels of his intelligence." Nowhere is this more apparent to me than in the video of Suda, a 4-year-old elephant from Thailand made to paint an elephant holding a red thorny flower with his trunk and the name "Suda" above it. The caption reads how elephants are one of the few species on this planet that are sentient and most comments are of the type "Wow, that's amazing, incredible or beautiful". And yet it is a sentient species taught to reproduce an image it may or may not understand for sake of pleasing crowds of species that has captured it and forced it to reproduce an image it may or may not understand. If that is not abrogating - repealing or doing away with - the use of certain higher levels of a sentient species, I don't know what is. As a poetic twist, one can imagine Suda painting an elephant on its hind heels with a "NO!" above it and then stampeding into freedom. But enough of getting sidetracked.

Communication about relationship (again)
Here he expands on the point that dolphin communication system is of an almost totally unfamiliar kind. Because we are mammals with large brains Bateson assumes that we are "familiar with, though largely unconscious" of "communicating about our relationships." He goes on to name different aspects of nonverbal communication that we share with other terrestial mammals: bodily movements, involuntary tension of voluntary muscles, changes of facial expression, hesitations, shifts in tempo of speech or movement, overtones of the voice, and irregularities of respiration.
He then takes example from Darwin's Expression... and says that if we want to know what a dog's bark "means" we should look at the "expressive" parts of his body. Darwin described a dog "in a savage or hostile frame of mind" walking upright and stiffly, with head slighly raised, tail erect and rigid, hairs along the neck and back bristling [standing upright], pricked ears directed forwards and eyes having a fixed stare (Darwin 1873: 50). Bateson repeats some of these items and goes on to elaborate his own extension of how sense of organs become "organs for the transmission of messages about relationship."
This account is relevant when compared to an aspect of George Herbert Mead's work. What Darwin calls "frame of mind" (especially in connection with communication about relationship) Mead, as a social behaviourist, calls "the motor attitude of a form [of organism] in its relation to others" - this comes to the fore in his treatment of how "Most social stimulation is found in the beginnings or early stages of social acts which serve as stimuli to other forms whom these acts would affect" (Mead 1912: 402). That is, "the baring of the teeth or lifting of the nostrils" have a functional social value as "early indications of an incipient act" (ibid, 402).
Compare this to Bateson's discussion of how the expressive parts of the body "tell you at what object of the environment he [the dog] barking, and what patterns of relationship to that object he is likely to follow in the next few seconds." Bateson suggests looking at the animal's sense organs: eyes, ears and nose. What he is hinting at is almost like the phenomenological contention about consciousness always being intentional towards an object. In this case, it is easily observable that what grasps the animal's attention is simultaneously what he is looking at, pricking hear ears or pointing his nose towards, etc.
An example from the human sphere is also put forth: a blind man "makes us uncomfortable, not because he cannot see" but "because he does not transmit to us through the movements of his eyes the messages we expect and need so that we may know and be sure of the state of our relationship to him." In studies of nonverbal communication this is referred to as orientation: "A person's orientation to another (e.g. facing towards or away) is as important or 'communicative' as the exact words being spoken (e.g. 'I really think a lot of you'), and the significance of each is partially dependent on the other" (Harper et al. 1978: 121-122).
An earlier and broader account of this phenomena found that "the orientation of organisms towards objects outside of themselves" leads us "to find significance in 'topistic' responses at the lowest biological level", such as even the sunflower turning towards the sun (Latif 1934: 59). In case of humans someone (I couldn't find a reference; Goffman maybe?) even suggested that there is such a thing as an "orientation response" that makes us orient towards or away from other people. E.g. you're walking on the street and when a likeable person passes you cannot help instinctively stepping towards them or by the same token navigating "around" a dislikeable or dangerous-looking passer-by.
In any case, "orientation" - whether postural or that of eyes - unavoidably transmits messages (Ekman and Friesen, 1969b: 60, would call these intrinsically coded messages, for the very fact of their occurrence has meaning) about the state of relationship. Bateson surmises that we won't know much about dolphin communication until we know whether dolphins are aware of these kinds of messages and can read them "in another's use, direction, volume, and pitch of echolocation". That is, the programmatic inquiry about those very specific signals about mutual awareness postulated in Communication... in 1951 still hold - if we want to gauge the nature of dolphin communication system then communication about relationship should serve as a starting point.
Next Bateson states that "Adaption to life in the ocean has stripped the whales of facial expression." On this point it is interesting to note Hediger's observation that the most important facial nerve in human beings, the nervus facilis that moves the right side of our face (and which is affected by Bell's palsy) "controls a considerable region, that of the gill-govers" in fishes; and it is wrong to regard underwater creatures as expressionless (Hediger 1968: 148). Though Bateson is probably correct in that "the conditions of life in the sea are such that even if a dolphin had a mobile face, the details of his expression would be visible to other dolphins only at rather short range, even in the clearest waters."
Thus he proposes that dolphin communication has gone through "an evolutionary shift from kinesics to vocalization" and ponders on how mammalian vocalizations are not totally opaque for us. Because we use grunts and groans, laughter and sobbing, modulations of breath while speaking and so on ourselves, we "learn rather easily to recognize" in animal vocalizations "certain kinds of greeting, pathos, rage, persuasion, and territoriality, though our guesses may often be wrong." The situation is sadly not the same with dolphins, whose vocalizations are, to repeat once again, "almost certainly of a totally unfamiliar kind".
Bateson distrusts the hunch that dolphin sounds are merely an elaboration of mammalian vocalizations and argues that he personally does not believe that dolphins have anything that a human linguist would call a language: "I do not think that any animal without hands would be stupid enough to arrive at so outlandish a mode of communication". What he does believe is that, like ourselves and other mammals, dolphins "are preoccupied with the patterns of their relationship". The last paragraph in this subheading goes into some details that I would like to elaborate with a discussion of communication models.

The μ function of the message
For this excursion I'd like to turn to Jurgen Ruesch's paper "Synopsis of the Theory of Human Communication" (1953) originally published in Psychiatry 16(3): 215-243, and republished in his Semiotic Approaches to Human Relations (Ruesch 1972 [1953]: 47-93). These titles by themselves should suggest that communication and relationship are intimately intertwined in his work.
Firstly, what was above discussed in vague terms like social acts serving "as stimuli to other forms whom these acts would affect" (Mead) and telling you "what patterns of relationship [...] is likely to follow in the next few seconds" (Bateson) is captured concisely in Ruesch's notion of metacommunicative instructions. E.g., an explicit metacommunicative instruction would occur when "a person who enters a room and introduces himself as the telephone repair man instructs the other people about his forthcoming actions"; and less explicit instructions are given through trough the uniforms of policemen, judges, and other officials (Ruesch 1972 [1953]: 76-77). But this is just a cursory similarity. Ruesch has a terribly thorough account of metacommunication, its functions, shadings, cultural implications etc.
There is hardly a more thorough overview of communication theory available anywhere anytime than is contained in Ruesch's "Synopsis..." and this is proven especially poignantly in the conclusion wherein Ruesch puts forth sever questions that the scientific observer may raise in the assessment of a communication system. The first six of these are are pretty much those of Roman Jakobson's in his famous paper "Linguistics and Poetics" (Jakobson 1981 [1958]). All but the last one, to which I will get just after the following exposition - when comparing Ruesch and Jakobson in this regard the six factors go as follows: (1) What are the limitations of the context? (Jakobson's context); (2) Who is saying it? (addresser); (3) To whom is it said? (addressee); (4) What is said? (message); (5) What media of communication are used? (channel); and (6) How it is being said? (code).
Although they differ greatly in the details of these aspects, the general model is almost exactly the same. Jakobson considers all but the very last of Ruesch's questions: (7) What is the result of the exchange of messages? This is the pivotal point I'd like to discuss. But before doing so I should note that Thomas Sebeok, in his paper "Semiotics and Ethology", puts forth a hexagonal model in which the very first element, "the source", is elaborated as bringing about "a relatively large redistribution of energy or matter in another animal (or in another part of the same animal)" (Sebeok 1969: 201). So not only does he address Ruesch's 7th question but without skipping a beat also makes room for autocommunication.
But we should also look at the specifics of Ruesch's 7th question. There are two major aspects about it, and we have to remember that his questions are geared towards assessing communication systems. The first concerns the correction of information at the source and destination of the message (the communication situation is viewed as a self-correcting system - there should be feedback to correct misunderstandings). The second concerns "the effects which [the communicative] action has had upon the existing communication system" (Ruesch 1972 [1953]: 89), or in Sebeok's words, what kind of redistribution of energy or matter the communication effected.
It is not enough to merely point out that the result of communication is missing from Jakobson's model. I have an inkling that this wasn't relevant for his basically "unilateral" (one-directional) model of communication because (a) his model is more like a typology of communication functions than an operational model for studying communication; and (b) it may very well be that he was primarily considering poetry as a form of communication when he first formulated it. I won't go into the details of Jakobson's model because it would take up too many pages if I did so, but for a while now I've been wondering how to "fill this gap", or what to place as the 7th element in an improved model. I also won't go into my first guess, because there's no need to. At this point I'd like to try out Bateson's "μ function of the message" as the 7th element.
The connection between the μ function and the result of communication may not be readily apparent. In fact it seems that some pretty difficult mental gymnastics are in order. Luckily, Ruesch himself can help us in this regard, as he is basically the co-author of the μ function. But first we should elucidate further what exactly characterizes the μ function according to Bateson: "Preverbal mammals communicate about things, when they must, by using what are primarily μ-function signals" such as the cat's mewing, translated as "Dependency!", that communicates "in terms of patterns and contingencies of relationship".
Turning to Ruesch, we find in his "Synopsis..." that "All human relations and communication systems are governed by rules; these are either handed down from generation to generation or newly created by mutual consent or by forceful imposition" (Ruesch 1972 [1953]: 62). We have already met something like this in Bateson when he discussed methodological considerations of experimentation. In the situation of the experiment an analogy can be drawn with humans having a handed-down definition of the rules (we have a general idea of how to cooperate within the experiment), while animal subjects don't have such luxury and have to be taught by "forceful imposition". This does not seem to get us closer to communication about relationship.
Turning to their co-authored Communication... we find a two-paragraph sub-chapter titled "The Perception of the Perception" that gives a more illustrative account. Here they begin by stating that "A social situation is established as soon as an exchange of communication takes place; and such exchange begins with the moment in which the actions of the other individual are perceived as responses - that is, as evoked by the sender's message and therefore as comments upon that message, giving the sender an opportunity of judging what the message meant to the receiver" (Ruesch & Bateson 1951: 23). It would be simple enough to call this feedback and be done with it, but that would be erroneous. What we have here is exactly the 7th element, or at least the "correction of information" aspect of it. Namely, the sender can perceive if the receiver understood the information contained in the message correctly and elaborate if necessary.
Next they affirm that this is a kind of "communication about communication" and necessary if an exchange of messages is to take place (that is, if communication in the truest sense of the word, of "making common", can occur). They call this phenomenon "the perception of the perception" and state that it is "the sign that a silent agreement has been reached by the participants, to the effect that mutual influence is to be expected" (ibid, 23). That is, a system of communication, as they call it, is contingent upon mutual awareness and influence.
Next come the illustrations: if person A raises his voice to attract person B's attention then "he is thereby making a statement about communication" (ibid, 24). Compared to Bateson's illustration of the mewing cat this raises a question: does the cat's raising her voice to attract her owner's attention make a statement about communication or about relationship? It seems indeed to be the case that no sharp line can be drawn between them. Another solution would be to equate the person's verbal statement and the cat's mew with "communication" and raising the voice to attract another persons attention and the specific tone, pitch or loudness of the mew with "communication about communication".
But this alternative solution does not lead us closer to understanding communication about relationship because it would isolate the communicative message from the communication situation. In this regard we must subscribe to their systemic approach and view communication as a system that is contingent upon mutual awareness and influence. Further illustrations seem to emphasize this point: statements like "I am communicating with you" and "I am not listening to you; I am doing the talking" acknowledge the very act or nature of the act of communication itself.
Thus it must be kept in mind that when they talk about communication about relationship they don't necessarily mean "relationships" in a lasting sense, but rather the fact or state of people (or other organisms) being connected through mutual awareness and influence in the given communication situation. Then again the illustration of pressing down the head as asserting or affirming the nature of the relationship between the wolf pack leader and the other male would argue against this. A more concise formulation can be found from the very beginning of the text, from the part about deutero-learning: "a relationship between two (or more) organisms is, in fact, a sequence of S-R sequences". We could now very well substitute "a sequence of S-R sequences" with a lengthier but more exact statement, "an exchange of messages in a situation of mutual awareness and influence".
Returning to the μ function of messages in Bateson's "Problems..." we can now look differently at his passing remark that "I ask for your attention and perhaps respect by talking about whales." This illustration works because it is communication about communication ("[I am] talking [to you] about whales"] as well as communication about relationship ("I ask for your attention and perhaps respect"). Thus far we have achieved some clearness, but there is another illustration in Communication... that finally forces me to discuss Jakobson's communication model.

The emotive function
The last illustration under "The Perception of the Perception" goes as follows: "If "A" adds the word "please" to a verbal request, he is making a statement about that request; he is giving instructions about the mood or role which he desires the listener to adopt when he interprets the verbal stream" (Ruesch & Bateson 1951: 24). It is clear that this added signal, the magic word "please", is a μ function message. What is significant about it is it's resemblance to Jakobson's emotive function. Before moving on to discussing the latter, I have to anchor down the sentence that followed the last one: "He is adding a signal to cause a modification in the receiver's interpretation" (ibid, 24). This part opens up a distinct possibility of going for a third lengthy excursion in this series, to Charles Morris's modors (cf. Morris 1949: 161). This text is already lengthy beyond courtesy so I don't see why not.
But first, the emotive function, which should be talked about exactly as emotive function: Jakobson himself adds "or 'expressive" function" with "expressive" between quotation marks as if to emphasize that this is an approximation. The emotive function is "focused on the addresser", "aims a direct expression of the speaker's attitude toward what he is speaking about" and "tends to produce an impression of a certain emotion, whether true or false"; with the last statement followed by the conclusion that "therefore, the term 'emotive,' launched and advocated by [Anton] Marty, has proved to be preferable to 'emotional'" (Jakobson 1981 [1958]: 22).
It is immediately clear that Jakobson's emotive function is readily comparable to Ruesch's metacommunication or Bateson's μ function. All three purport to communication about communication. For example, the above illustration of adding the word "please" to a verbal request can be said to express the speaker's attitude towards his request (it is meant politely; or meant to as-if coax the listener to adopt a role), or it tends to produce an impression of a certain emotion - in Ruesch & Bateson's formulation, it gives instruction about the mood in which the speaker desires the listener to interpret the request.
I claimed above that Jakobson's "communication model" is not so much a communication model as a typology of different functions of language. He himself calls his "model" merely a "schematization". The typological aim is laid bare in this case with the sentence, "The purely emotive stratum in language is presented by the interjections." (ibid, 22). It is an error to think only of interjections as emotive, as he insightfully adds that the emotive function "flavors to some extent all our utterances" (ibid, 22). For sake of staying on animal communication even a bit we could probably compare the cat's mew as an interjection, although there is no real necessity for doing so.
It is now time to turn to our unknown authority on emotive function, the mysterious Anton Marty. As it turns out he was a Swiss philosopher of language and a rector of the University of Prague. He died in 1914 but his ideas were positively received by the Prague Linguistic Circle that arose after his death, perhaps even decisive influence, for he argued "that language is functional and therefore different criteria have to be applied when language is used for communication as opposed to when used in poetry" (Lečka 1995: 86-87; in Nylund 2013: 195). Here we see a differentiation between artistic and non-artistic use of language, or between language with poetic value and without... But I also see that I'm getting sidetracked. At this point it is only somewhat relevant to mention that "Vilém Mathesius, Roman Jakobson, Bohumil Trnka and Jan Mukařovský were all influenced by Marty's ideas" (ibid, 195) and the melting pot of ideas that was the Prague Linguistic Circle undoubtedly influenced Jakobson's scheme of different functions of language.
Since Marty's Untersuchungen zur Grundlegung der allgemeinen Grammatik und Sprachphilosophie, Vol. 1 (1908) has not been translated into English, I can only rely on Standford Encyclopedia of Philosophy's abstract on the matter. Moreover, before I can do that, I have to point out that Marty was a prominent student of Franz Brentano. This is relevant only because Marty's emotives were influenced by Brentano's odd theories about how emotions are mental phenomena and thus intentional (cf. Mulligan 2004).
For our purposes it is not necessary to go into more nitty-gritty of Brentano's thought than is necessary for understanding emotives. Thus I'll limit myself to a quotation that "autosemantic signs are divided into 'emotives' (exclamatory sentences, commands, questions), 'assertions' (declarative sentences) and 'suggestives of representations', i.e. nouns" (Marty 1908: 476; in Graffi 2001: 62) and that this classification is based on Brentano's taxonomy of psychic phenomena, so that emotives are based specifically on Brentano's Gemütsbewegungen (ibid, 62).
Now we can arrive at the Stanford Encyclopedia abstract where it is written that as "Marty regards statements as autosemantica which manifest judgments and communicate to the interlocutor that he or she is to judge in the same way, he characterizes emotives or interest-demanding expressions as those autosemantica which manifest not only emotions, but also volitions, and communicate to the interlocutor that he or she is to feel or will in the same way." That is, we arrive at a point where we can revisit Jakobson's statement about the emotive function tending to produce an impression of a certain emotion and re-interpret it to the effect that the emotive function intends to produce in the receiver the impression of a certain emotion.
What does this small difference in wording amount to? Basically it leads us back to where we already were: the emotive flavor of all our utterances is added intentionally by the sender "to cause a modification in the receiver's interpretation", as Ruesch and Bateson put it above (1951: 24). That is, the "please" in a verbal request not only tends to produce a certain mood in the listener but is in fact intended as such.
To tie up this specific excursion and move on to the next one (which will hopefully be shorter), I have to modify my original assumption that Ruesch's 7th question is completely missing from Jakobson's communication model. When looked at in detail it turns out that the question of the result of the exchange of messages is merely not explicit in Jakobson - it is in fact hidden in something like a footnote that most readers just pass by; it is hidden in a reference to Anton Marty.

Modors
I do intend to return to reviewing Bateson's "Problems...", but as these last excursions have been interesting in themselves and do link up with the latter parts of Bateson's paper, I see no reason why I can't discuss another perspective on these problems. The tangible connection with Charles Morris's category of signs called modors appears in the the statement that when someone adds the word "please" to a verbal request, "He is adding a signal to cause a modification in the receiver's interpretation" (Ruesch & Bateson 1951: 24).
Modors appear in Morris's Signs, Language and Behavior (1949) in a sub-chapter titled "Kinds of Formators". The latter "serve to influence in a particular sign combination the way in which certain component signs signify or the way in which the combination as a whole signifies" (1949: 159). For example, the word "please" in a request influences the way in which the request as a whole signifies.
In his almost omnipotent prowess for technical distinctions, Morris outlines three classes of formators. Although a good case could be made for other classes of formators (determinors and connectors) operating in metacommunication, in the μ function or even in relation with the emotive function, my own technical prowess is not comparable to that of Morris, so I am limited to pointing out the connection between the former terms and only one class of formators: the modors. And even here my own contribution is negligible.
At the core of the matter, we can most readily rely on Morris's classification of intonations and speech melodies as modors (1949: 161). From here on out there is no getting around Morris's own highly technical jargon. Modors "are formators which establish the ascriptive mode of signifying of the sign combination in which they occur" (ibid, 161). That is, the modor determines the type of ascriptive mode of signifying of the sign combination in which they occur. The ascriptive mode of signifying can be either designative (inform), appraisive (provoke valuations) or prescriptive (incite actions in both linguistic and non-linguistic matters).
Morris's sharpened-arrows-like technical terms are almost incomprehensible without illustrations. In our above illustration, the "please" is a modor and the request itself is the sign combination in which it occurs. Because we already know that the illustration is a request we can unhesitantly classify it as a prescriptive ascriptor. Morris's own illustration is based on the sentence "He is coming". This sentence can be spoken in different ways: in the ascriptive mode of signifying it would be a mere statement that informs listeners that this event is indeed occurring; in the appraisive mode of signifying "He is coming" would signify that his coming is good or bad (e.g. "Oh no, he is coming!" or "Oh yes, he is finally coming!); in the prescriptive mode of signifying the statement incites an inquiry ("Is he really coming?") or ascribes a demand ("He must come!").
My verbal approximations are not the best way to illustrate these modes, as the actual difference is paralinguistic: designative would be neutral fact-stating; appraisive would contain a value-judgment whether the idea of him coming is accepted or rejected; and prescriptive would be self-assured utterance of the type "it must be so". Morris adds that in writing, modors correspond to punctuation devices like ".", "!", "?" and "!!", and notes that there is no sharp standardization in this regard. Today we could probably call emoticons modors. Moreover, one could approach most all nonverbal communication that accompanies verbal communication as modors, if only one could stick to Morris's technical terminology that goes along with it.
In the end, modors are indeed a form of communication about communication and communication about relationship, although elucidating the latter aspect would necessitate going into more technical details about modes and types of signifying. That is, just like metacommunication, the μ function and the emotive function, modors modify the interpretation of a given message. The difficulty in applying it comes to the fore when we consider the illustration of the cat's mew, for example, in which case we would somehow have to draw a sharp line between the mew as a vocalization pure and simple and it's intonation. There's also the issue of modors being somehow signs that ascribe the mode of signifying of sign combinations other than itself. Although Morris's classification is extremely sophisticated, it is perhaps too sophisticated for its own good.

Analogic versus digital communication (again?)
Bateson questions how it happens that "the paralinguistic and kinesics of men from strange cultures, and even the paralinguistics of other terrestial mammals, are at least partially intelligible to us, whereas the verbal language of men from strange cultures seem to be totally opaque?" A simple answer would refer to the anatomical and physiological buildup in men from strange cultures and partly even of other terrestial mammals. Nonverbal communication has a much broader evolutionary history than even the mere faculty for verbal communication. But then again he's not trying to ascertain an answer to this question but refers to it in order to point out that dolphin vocalizations are more like human language than the kinesics or paralinguistics of terrestial animals. That is, they are, once again, "of a totally unfamiliar kind".
He explains this unfamiliarity in terms of digital (language) and analogic (kinesics and paralinguistics) codifications, although he doesn't quite explain yet exactly why "gestures and tones of voice are partly intelligbile while foreign languages are unintelligible". We'll have to anchor this connection down and return to it after Bateson has made a case for these terms.
We begin our process of understanding these terms with digital communication, in which a number of purely conventional signs are "pushed around according to rules called algorithms. He says that digital signs "have no simple connection (e.g., correspondence of magnitude) with what they stand for". We'll have to consider the "simplicity" of a correspondence of magnitude when we reach analogic codification. Here he contrasts analogic "magnitude" with digital signs through the illustration that the numeral "5" is not (visually) bigger than the numeral "3" - what makes us think that the number 5 is bigger than 3 is pure convention. This convention is best illustrated by a name, as names have a purely arbitrary connection with what class the name stands for. The numeral "5" is only the name of a magnitude. He goes on to demonstrate this by noting that "It is nonsense to ask if my phone number is larger than yours," because the telephone system is digital.
"In analogic communication," he continues, "real magnitudes are used, and they correspond to real magnitudes in the subject of discourse." His example of an "analogue computer" is the rangefinder camera. Since this technology is long outdated and modern readers have little grasp of what the issue is about, we can probably skip this illustration. He goes on to apply the same absurd numerical example on words: "The word 'big" is not bigger than the word 'little'", and so on. The only valuable piece of information here seems to be the suggestion that in analogic codification the sign corresponds "to the system of interrelated magnitudes in the object denoted."
The significance of this latter passage consists in the nature of the interrelation between the analogic sign and its denotation. His examples include the magnitude of the gesture, the loudness of the voice, the length of the pause, the tension of the muscle, etc. These outward - even measurable - signs supposedly "correspond (directly or indirectly) to magnitudes in the relationship that is the subject of discourse". Thus the "object denoted" accords primarily to the "real magnitudes in the subject of discourse". This relationship has very definite implications for understanding analogic communication.
...
Now at this point I'll have to stop this essay. I won't even give it a proper conclusion because in my mind it is not finished. I can return to writing this essay when I have dealt more with Jakobson and his view of communication. I believe that looking at Jakobson's more strictly "digital" (linguistic) understanding of communication will ultimately prove the valuable contribution of Bateson's view. But that's for another day and another post.

Presentation

• Essay on Bateson's "Problems..."
• Video clips
  1. AN ECOLOGY OF MIND (Bullfrog Films clip)
  2. 6 Scientific Findings That Prove Cats Are Evil