Hi Theorists,

Following on from the discussions of multiple ontologies and perspectivism, I'm trying to build an argument about the incommensurability of ontologies (in plain language, how people's assumptions about the nature of reality prevent them - at least at times - from getting through to each other, e.g. as observable ethnographically in 'first contact' situations, or indeed in much anthropological fieldwork in general).

In particular, I'm interested in how people actually go about forging intelligibility between each others 'worlds' - in whether and to what degree they are able to learn each others' 'ways of seeing' (ontologically speaking) - and what the 'mechanisms' might be for doing that.

In other words, how do people with different assumptions about the nature of reality get on (where and when they do), and how do ontologies change?

Aside from my own ethnography, Quine seems useful, but on the other hand there must be some basic reasons as to why he doesn't seem to feature very prominently in anthropological discussions, despite his own interest in the subject (gavagai's etc.)

So here I'm curious to learn, from your perspectives, what's wrong with Quine, as far as anthropology is concerned?

Views: 1251

Replies to This Discussion

Hi Amy,

I'm dashing, I'll try to be brief and deal with Quine only. So far the only anthropologist I know of who dealt with Quine is Layton. He briefly discusses Quine in his theory inthro book, p. 199, in the intro in his ed volume Arch and Anth of landscape and his article in After Writing culture. These are the only ones I remember right now. L. points out the classic Quinean discussion of the gavagai problem in Words and Object of the anthropologists watching the native pointing towards a rabbit and uttering the word 'gavagai'. I remember that Quine argues that one may never know whether gavagai refers to the rabbit as being, to the grass or the rabbit-event. Layton follows Quine and try to differentiate between 'ostensive' statement addressed to external referents and theoretical and abstract statements and come to the conclusion that environments elicits meaning (very rough summary!).

The problem was addressed by Carrie in Shamans and Elders. Stepping out from representational paradigms she refers to the relation between processual aspects of external referents and signification (first chapter on landscape). For example, how in Mongolia mountains 'stand' for the solidity of clans or water and rivers stand for change. Eco deals with the issue referring to Pierce's notion of Ground of things in Kant and the Platypus. In other words, there are certain 'structural' constraints in referents which leads signification in certain directions (i.e. why the snake is employed to 'represent' cyclical movement instead than the hippo) It's been Eco's obsession in several of his works (Limits of Interpretation, Treaty of General Semiotics). I actually remember he wrote a book on translation where he deals with the problem of translating different ontologies, the English title is Experiences in Translation, I think that's what you are looking. Carrie also wrote an article which includes several of the discussions we had in the last years in relation to the idea of 'individual subjects' and events (Reassembling Individual Subjects in Anthropological Theory 2008, just google it)

The only thing which may be wrong in Quine when he presents the gavagai problem is that in order to work his anthropological vignette must be 'purified' (in the Latourian sense). An anthropologist would never find himself in the gavagai situation but will have other clues, before and after the 'event', to frame the meaning of the native's utterance.

About your general questions, I think the answers are not easy and time would require time I don't have right now!

All bests,

Giovanni
Could it be that there is nothing wrong with Quine, except perhaps for the fact that few anthropologists have read him. My encounter with him was in the early 1960s, when I was an undergraduate taking a course on symbolic logic, using Methods in Logic as a text. It was decades later when I heard it rumored that he, along with Davidson, marked a linguistic turn in American pragmatism, which had previous, in its James, Pierce, Dewey version focused on experience instead of language. But even now I haven't found the time to go back and read Word and Object seriously.

It might be interesting to take a quick poll and ask how many people here

1. Have heard of Quine
2. Know that he was a prominent American philosopher in the 1950s and 60s
3. Have a clue what he was talking about.
4. Actually have an informed opinion about what he wrote.
The most overt use of Quine (and Davidson) I know of in anthropology is in Elizabeth Povinelli's Cunning of Recognition. It's very very clever.

While I think John is right that one of the reasons he hasn't found much of a home in anthropology is because he simply isn't read-- and moreover anthropologists aren't trained *how* to read analytic philosophy and often find it very difficult-- I think another important reason is that Quine is committed to physicalism and extensionalism, and we are generally speaking not.
Thanks Giuseppe and Giovanni, you've both offered really helpful responses to my question.

Giuseppe, you're absolutely right that power can't be removed from the mix - but I want to approach these issues from a slightly different direction, just for the moment (though I look forward very much to hearing more about your work).

I come to this from an interest in what the artefacts of early contact exchanges might tell us about how incommensurability gets worked through in practice (in other words what happens when 'worlds collide', as it is often put, or how people approach one another through things). This, I hope, will tell me something about how people build intelligibility when they share no common ground, and how (whether?) one's most basic assumptions about life are subject to change.

My interest is thus not primarily in 'symbolic manipulations', but rather in how people get through to each other (if and when they do) and manage to carry out joint projects (again if and when) even before they share symbolic vocabulary of any kind (linguistic or otherwise). 'First contacts' are interesting because they seem (from one perspective at least) to throw the gritty, 'embodied' reality of incommensurability into stark relief (indeed, unintelligibility often leads to killings - a most effective way of excluding others from your world!). There are endless examples of incomprehension and misunderstanding, even where supposedly straightforward ostension is involved (e.g. pointing at something to say that you want it). Yet pretty often people do manage to exchange things (including names, food, bodily fluids and parts of their identity) and do stuff together (navigation, sex) to more or less mutual satisfaction. Pretty quickly, of course, language comes into play, but in a very haphazard and hit-and-miss sort of way - and it stays like that for a long time, sometimes over generations.

What I found useful about Quine is that, although he's most famous for pointing this out with the gavagai example (one focused on the problem of radical incommensurability) he also suggested that such hit-and-miss impasses - moments of unintelligibility - are a feature of all communication, with his statement that 'radical translation begins at home'. (So, Giovanni, even where the anthropologist has access to interpretive clues, "before and after the 'event'", these are no guarantee of success). I found Quine's explanation of this pretty tough-going (it involves a complex disquisition on the inscrutability of reference, and the problems both of ‘direct’ and ‘deferred’ ostension (Ontological Relativity and other Essays 1969: 38-48). But what I take from it is roughly that ontology develops through experience (e.g. babies learn to objectify - where they do - from their mothers), and since people's experiences differ, so do their 'ontologies' (ontology is not, therefore, just another word for culture).

To me this opens the possibility of thinking about 'ontological' dynamism and interaction (two issues that don't seem to feature much in the 'multiple ontologies' discussions) in interesting ways. Particularly, it affords possibilities for thinking through these problems, that were foreclosed by all versions of the notion of 'culture' that I can think of. If ontologies must be shared, as John's Wikipedia definition of 'ontology' claims (cited in the earlier discussion), then they are not a property of individuals. Moreover, if they emerge from one's engagements with 'others' (in the broadest sense - not just other people), then one presumably shares different bits of one's ontology (or different ontologies) with different people - so nor can 'ontology' be mapped neatly onto 'culture'.

Giovanni, thanks for the brilliant bibliographic references, which I will certainly follow up on - just what I needed!

To be a bit more specific about what I think might be one of the problems with Quine, I know that although he was interested in ontological multiplicity as a problem of translation, he remained committed to his own ‘one-world’ ontological position: ‘Our physics is provincial only in that there is no universal basis for translating it into remote languages; it would still never condone defining physical identity in terms of verbal behaviour. If we rest the identity of attributes on an admittedly local relation of English synonymy, then we count attributes secondary to language in a way that physical objects are not’ (1969: 20). This is hardly in line with the program, as far as the 'multiple ontologies' school is concerned! His disclaimer at the end of ‘Speaking of Objects’ may be read as a bit of a cop-out, ethnographically speaking: ‘In saying this I philosophize from the vantage point only of our own provincial conceptual scheme and scientific epoch, true; but I know no better’.

Amiria
Thanks Adam (for this and the slightly more detailed version you've posted over on 'Linguistic Anthropology').

If it's not asking too much, would you be able, for the benefit of us non-philosophers, to expand a little on what you referred to as Quine's "commitment to an extensionalist account of meaning, which we [linguistic anthropologists?] certainly do not share".

Also what is meant by his "physicalism". What are (some of) the important premises that have been abandoned?

I will definitely check out Povinelli.
Cheers, Amiria

Adam Leeds said:
The most overt use of Quine (and Davidson) I know of in anthropology is in Elizabeth Povinelli's Cunning of Recognition. It's very very clever.

While I think John is right that one of the reasons he hasn't found much of a home in anthropology is because he simply isn't read-- and moreover anthropologists aren't trained *how* to read analytic philosophy and often find it very difficult-- I think another important reason is that Quine is committed to physicalism and extensionalism, and we are generally speaking not.
So, you cited a quote attesting to his physicalism just above. Quine thinks physical science is the one true theory of what is out there. What *really* exists are the entities required by our best physical theory; everything else should be reducible to them or is meaningless. With respect to his beliefs about language, Quine's philosophy of language is in the framework of a future purely behavioral psychology. He wants to talk about dispositions, stimuli, etc. Meanings, however, are not to be part of the theory of language. Talk of "understanding" as meaning anything more than dispositions to indicate assent to similar stimuli is otiose to him.

Very concretely, the arguments for the indeterminacy of translation are not arguments that at first, or sometimes, we don't understand each other. Indeterminacy of translation is a skeptical argument about the very existence of semantic facts. According to this argument, there is simply no fact of the matter about linguistic expressions mean. Is such an argument useful to anthropologists? I couldn't see how. However, this argument requires Quine's assumption of very thin behavioristic evidence as the only admissible evidence for beginning attempts to compose a translation manual. Many of us, again, would find this limitation implausible. It is unclear whether Quine's argument can be separated from these premises.

This was pretty compressed. Was it helpful?

Unrelated to the above, you're going to also want to look at Povinelli's review article on The Anthropology of Incommensurability in the Annual Review of 2001.

Amiria Salmond said:
Thanks Adam (for this and the slightly more detailed version you've posted over on 'Linguistic Anthropology').

If it's not asking too much, would you be able, for the benefit of us non-philosophers, to expand a little on what you referred to as Quine's "commitment to an extensionalist account of meaning, which we [linguistic anthropologists?] certainly do not share".

Also what is meant by his "physicalism". What are (some of) the important premises that have been abandoned?

I will definitely check out Povinelli.
Cheers, Amiria

Adam Leeds said:
The most overt use of Quine (and Davidson) I know of in anthropology is in Elizabeth Povinelli's Cunning of Recognition. It's very very clever.

While I think John is right that one of the reasons he hasn't found much of a home in anthropology is because he simply isn't read-- and moreover anthropologists aren't trained *how* to read analytic philosophy and often find it very difficult-- I think another important reason is that Quine is committed to physicalism and extensionalism, and we are generally speaking not.
Thanks Adam,

This is indeed very helpful.

AL (channeling Quine): "Meanings, however, are not to be part of the theory of language."

Well, if meanings have to be something separate from their referents, then this might start looking like a good idea - but perhaps this isn't the point...?

AL: "Indeterminacy of translation is a skeptical argument about the very existence of semantic facts. According to this argument, there is simply no fact of the matter about linguistic expressions mean. Is such an argument useful to anthropologists? I couldn't see how."

Well, if one of the aims of anthropology is to take our informants seriously (i.e. not automatically assume they are speaking metaphorically when they say that 'twins are birds' or that they simply 'represent the world differently'), then getting rid of semantic facts might help - or not? Could you elaborate on what is meant by 'semantic facts'?

AL: "However, this argument requires Quine's assumption of very thin behavioristic evidence as the only admissible evidence for beginning attempts to compose a translation manual. Many of us, again, would find this limitation implausible. It is unclear whether Quine's argument can be separated from these premises."

Could you articulate a little more explicitly how this is a problem for anthropologists - why many of us "would find this limitation implausible"?

I think I've got the second Povinelli article you mention - will have another look at it in light of the discussion.

Thanks again,
Amiria
"Indeterminacy of translation is a skeptical argument about the very existence of semantic facts. According to this argument, there is simply no fact of the matter about linguistic expressions mean. Is such an argument useful to anthropologists? I couldn't see how."

Would it, I wonder, be useful to reflect on the way in which "fact" is used in these claims? As I mentioned before, it has been a long time since I read this stuff and I haven't read at all more recent scholarship, e.g., Povinelli. What I do seem to recall, however, is that "fact" at the time that Quine was writing was equivalent to a proposition that was demonstrably true or false plus sufficient evidence to support the claim that it was true, where evidence was sufficient if and only if it was empirical and collected using scientific methods.

If one looked into the history of this characterization of "fact," the starting point was Aristotle and the classic distinction between essence and accident in characterizing substances. Semantic facts were, for all practical purposes, essences. The essential assumption (pun intended) behind semantic analysis was that there is, for the term in question, a single proper meaning for the term describable in terms of necessary and sufficient conditions. Everything else—nuance, emotion, metaphorical usage, performative effect—was superfluous and not describable in terms of the necessary and sufficient conditions that determine a the fact, a.k.a., the single proper meaning.

Why is this sort of thing of anthropological interest? It interests this anthropologist because it is an example of what I call the rationalizing move, an intellectual gesture found in all sorts of places and traditions. The gesture occurs in situations when, for whatever reason, someone has to decide if x counts for purpose y. The someone can be a child attempting to understand the range of animals to which a word like "cat" applies or what she can still get away with when mother says "no." It can also be a Confucian or other traditional intellectual attempting sort out what is essential to the tradition, as opposed to "superstitious" or "feudal," or "irrational" accretions that can be dispensed with as part of modernizing local thought.
The basic issue is always the same, keep it or discard it. The anthropology comes in because what is kept and what is discarded varies widely and the attempt to understand how the decision is made is a deep and fascinating problem.

Just one example and then I'll shut up. In his study of a village in southern Taiwan, David Jordan observes that in his village young men sometimes fall into trance; they tremble, they shake, they seem unconscious of what they are doing. In the village there are four possible meanings that might be ascribed to this behavior: (1) the young man is possessed by a god, (2) the young man is possessed by a ghost, (3)the young man is sick, (4) the young man is faking. (1) is a good thing, because it is useful to have a spirit medium around through whom a god can be addressed, when the god's advice or assistance is needed. (2) is a bad thing but performing an exorcism makes the ghost go away. (3) is a bad thing, but exorcisms don't work and, even with training, the sort of response to queries expected of a god never appears. (4) is always a possibility, since acceptance as a medium involves prestige and a money-making opportunities that motivate charlatans. Sorting out the processes by which one or another interpretation becomes the "social fact" on which the village agrees. That's the deep and fascinating problem I mentioned—made even more so because there may or may not be important differences between what goes on in a village as opposed, for example, to an academic discipline.
Just a brief suggestion. Computer scientists and knowledge engineers frequently encounter the problem of having to translate or bridging between ontologies. The rise of semantic markup and the web has made this an even more pressing problem. For example, two documents might contain relevant information, but without some kind of appropriate thesaurus between them, it can be really difficult to compare the two documents' contents. This is a really difficult problem, especially in the relatively impoverished information context of such textual documents.

I think that the keyword is "ontological alignment" or "ontology mapping". Here are some references:

good overview of ontology alignment/mapping
Kalfoglou, Yannis, and Marco Schorlemmer. 2003. Ontology mapping: the state of the art. Knowl. Eng. Rev. 18, no. 1: 1-31.

excellent textbook on knowledge engineering (I'm reading it now)
Sowa, John F. 2000. Knowledge representation. Brooks/Cole.

(uses Jon Barwise's channel theory)
Schorlemmer, Marco, and Yannis Kalfoglou. 2005. Progressive ontology alignment for meaning coordination: an information-theoretic foundation. In Proceedings of the fourth international joint conference on Autonomous agents and multiagent systems, 737-744. The Netherlands: ACM. http://portal.acm.org.hmlproxy.lib.csufresno.edu/citation.cfm?id=10....

semantic web, SKOS Simple Knowledge Organization System
http://www.w3.org/2004/02/skos/

Engineers are often less constrained that the social scientist, in that any solution that works is more or less acceptable while an anthropologist or other social and behavioral scientist must (or ought to) base his or her description on empirical realities particular to our species. But engineer's solutions generally have to work, demonstrably; worth taking a look, if only to define the more general constraints of such problems, or to gain inspiration.
If, perhaps I might add that although I am almost entirely ignorant of the philosophy of Quine and whether my suggestion will be useful, but situation semantics and situation theory might be an interesting methodological framework in which to investigate such ontological semi-incommensurability. Situation semantics does have something of an unsettled and controversial status within philosophical semantics, but two things at least recommend it. First, it is based on the notion of information, rather than that of truth. Second it is fundamentally based on the notion of the context dependence of meaning. Situation semantics and situation theory are, however, a realist sort of semantics. That is hardly a mark against it in my book, but some might be more skeptical. I have the feeling that the notion of incommensurability may challenge this framework, but, I hope, in a constructive fashion. In any case, a useful introduction can be freely downloaded online:

Keith, Devlin. 2006. Situation Theory and Situation Semantics. In Handbook of the History of Logic, ed. John Woods and Dov Gabbay, 7:601-664. Elsevier. http://stanford.edu/~kdevlin/HHL_SituationTheory.pdf.

Sorry, that my last two posts have poured on the references.
So I've done a bit of reading overnight and it seems that Quine is many things to many people (in fact he has spawned an entire academic industry, for those who were wondering why I don't just go and read all the relevant secondary sources myself. Answer: there are simply too many of them than I could tackle in a lifetime!) Nonetheless, a few relevant points may be gleaned:

AL: "Very concretely, the arguments for the indeterminacy of translation are not arguments that at first, or sometimes, we don't understand each other."

I see what you mean, Adam, in the sense that Quine's concern here is more with establishing an empirical baseline for linguistic theory (determining which bits of language are available to science, empirically conceived) than with questions of 'understanding' per se. Indeed, some extrapolations of his argument suggest that 'true' understanding in the sense of a 'meeting of minds' or the 'transmission of [objective] meaning' is impossible, since (empirically significant) meaning is for him (insofar as it exists) an artefact of encounter(s) between people, no two of which are identical (see below).

At the same time, the point of all this (at least as far as anthropology is concerned) seems to be simply to clear away certain prevalent assumptions people have (had) about language, in order to crack on with the business of looking at how people actually get by in terms of getting stuff through to each other, no?

On semantic facts:

JC: "The essential assumption (pun intended) behind semantic analysis was that there is, for the term in question, a single proper meaning for the term describable in terms of necessary and sufficient conditions. Everything else—nuance, emotion, metaphorical usage, performative effect—was superfluous and not describable in terms of the necessary and sufficient conditions that determine a the fact, a.k.a., the single proper meaning."

AL: "Indeterminacy of translation is a skeptical argument about the very existence of semantic facts. According to this argument, there is simply no fact of the matter about linguistic expressions mean. Is such an argument useful to anthropologists? I couldn't see how."

As Adam notes, Quine 'got rid of' semantic facts (i.e. the idea that a particular sentence has an objective meaning). Funnily enough, this wasn't because he wanted to interfere with the idea that objective facts exist (as many anths might want to), but rather because, he argued, meaning conceived as a property of mind is incapable of being addressed empirically (except in it's socially observable manifestations) and therefore it didn't deserve to exist as an object of empirical science, in and of itself. So far so good for anthropologists, I reckon, since I don't think we need to have a stake in meaning as an objective fact, as long as we can have it in 'socially observable' forms (which would include a number of those mentioned by John above).

Quine (cited in Gary Kemp, Quine for the Perplexed (!): "Language is a social art. In acquiring it we have to depend entirely on intersubjectively available cues as to what to say and when. Hence there is no justification for collating linguistic meanings, unless in terms of men's dispositions to respond overtly to socially observable stimulations (WO ix; see also OR 26-9, WP 221-7)."

Kemp: "Quine is assuming that meaning must be public, in the following sense: if two speakers A and B were entirely and exactly alike in all their verbal dispositions - their dispositions to speak in various circumstances - then it could not be the case that for any sentence, what A means by it is different from what B means by it."

Obviously no two people are identical in their 'verbal dispositions' (or ontologies, one might add), which 'proves' that meaning independent of people's interactions doesn't exist (at least in the sense that its existence cannot be established empirically).

Kemp continues: "The reason for this in principle is quite clear. each of us learns to speak only by observing other people and learning to talk as they do - for example, to answer 'yes' or something similar when asked, in the evident presence of rain, 'Is it raining?' All we have to go on, in determining whether or not we understand other people - determine what they mean - is what they say in what situations. Of course one can remain sceptical: perhaps, despite the fact that communication seems to proceed smoothly, we can never really understand each other. Perhaps we only ever understand our own language, never that of anyone else! Wittgenstein, famously, argued that such private understanding or private language is actually impossible: understanding is necessarily something mutual, something corroborated in the public domain. Although Quine approves of the conclusion, he does not lean on anything like Wittgenstein's private language argument per se. Instead, Quine should be understood to be concerned with that ordinary concept of meaning that we use in ordinary life, whereby apparently smooth communication justifies ascriptions of meaning.... Something necessarily inscrutable in his mind would not be relevant to meaning, in this sense, even if there were or could be such a thing" (WP 221-7).

Sounds alright to me, as far as anthropology goes....
Thanks Jacob - these points are really interesting, and the references very welcome - especially because we'll be doing some actual 'ontological mapping' and 'ontological alignment' in the project I'm currently working on, about early European voyaging in the Pacific. This involves bringing various databases of digitized primary source material (objects, images and text) together into a 'digital research environment'. The challenges of managing diverse digital forms (sound and video files, images, text - marked up and otherwise) are obvious! I may get back to you on these issues later, if that's okay?

RSS

Translate

OAC Press

@OpenAnthCoop

Events

© 2019   Created by Keith Hart.   Powered by

Badges  |  Report an Issue  |  Terms of Service