New York Jewish Conversational Style.

Back in 2007 I posted about a radio talk Deborah Tannen gave about “New York Style”; now here’s her early paper on a version of the topic, “New York Jewish Conversational Style” (International Journal of the Sociology of Language, 30[1981]:133–149). That link is the paywalled De Gruyter one, but you can read the paper here, and an enjoyable read it is. Some excerpts:

My own findings on New York Jewish conversational style were in a way serendipitous as well. I had begun with the goal of discovering the features that made up the styles of each participant in two-and-a-half hours of naturally occurring conversation at dinner on Thanksgiving 1978. Analysis revealed, however, that three of the participants, all natives of New York of East European Jewish background, shared many stylistic features which could be seen to have a positive effect when used with each other and a negative effect when used with the three others. Moreover, the evening’s interaction was later characterized by three of the participants (independently) as “‘New York Jewish’ or ‘New York’”. Finally, whereas the tapes contained many examples of interchanges between two or three of the New Yorkers, it had no examples of talk among non-New Yorkers in which the New Yorkers did not participate. Thus, what began as a general study of conversational style ended by becoming an analysis of New York Jewish conversational style (Tannen, 1979).

“Style” is not something extra, added on like frosting on a cake. It is the stuff of which the linguistic cake is made: pitch, amplitude, intonation, voice quality, lexical and syntactic choice, rate of speech and turntaking, as well as what is said and how discourse cohesion is achieved. In other words, style refers to all the ways speakers encode meaning in language and convey how they intend their talk to be understood. Insofar as speakers from similar speech communities share such linguistic conventions, style is a social phenomenon. Insofar as speakers use particular features in particular combinations and in various settings, to that extent style is an individual phenomenon.

The emotional/aesthetic experience of a perfectly tuned conversation is as ecstatic as an artistic experience. The satisfaction of having communicated successfully goes beyond the pleasure of being understood in the narrow sense. It is a ratification of one’s place in the world and one’s way of being human. It is, as Becker calls a well-performed shadow play, ‘a vision of sanity’.

This raises the question of the extent to which the linguistic conventions I have discussed are ‘New York’ and/or ‘Jewish’. My hypothesis is that the style (i.e., the combination of linguistic devices used in the way described) I have discussed represents a prototype of a kind of conversation that is familiar to most New York Jews and unfamiliar to most midwestern and western Americans of non-Jewish background. My impression is that New Yorkers of non-Jewish background and Jews not from New York City use many of the devices I have described and that there are New York Jews who use few of them. I suspect that the existence of this style represents the influence of conversational norms of East European Jewish immigrants and that similar norms are probably general to the Levant. I have not encountered evidence to indicate that Jews of German background necessarily share this style.

I was brought up with a very different conversational style (your basic non-NYC Protestant politeness, with everyone taking turns), but as soon as I moved to the city I embraced the local ways with delight (along with the subways and bagels). Incidentally, I liked this section from Wikipedia’s article on Tannen:

Tannen analyzed the agonistic framing of academic texts, which are characterized by their “ritualized adversativeness”. She argued that expectations for academic papers in the US place the highest importance on presenting the weaknesses of an existing, opposing, argument as a basis for bolstering the author’s replacement argument. According to her, agonism limits the depth of arguments and learning, since authors who follow the convention pass up opportunities to acknowledge strengths in the texts they are arguing against; in addition, this places the newest, attention-grabbing works in prime positions to be torn apart.

This doesn’t just happen with academic papers — it seems to be the basic rule of online discourse these days. Don’t bother finding something to appreciate, just look for an error, or something you disagree with, and announce that that’s where you stopped reading, since it proves the author is a moron. Me, I assume we all are wrong much of the time, but most of us have something to contribute anyway, and that’s what I’m interested in.


  1. Stu Clayton says

    “Style” is not something extra, added on like frosting on a cake. It is the stuff of which the linguistic cake is made

    So style is the batter. But then where does meaning fit in ? Is it the frosting ?!

  2. PlasticPaddy says

    I interpreted her point as being that the (intended) meaning is not (or not always) directly derivable from the words but moderated by the style, which then is like a binding agent, giving the cake its shape. The (intended/perceived) meaning is then the (intended/perceived) visual impression. Or if you want to go with taste, the style is the added sugar, salt, moistener etc., used to moderate the taste of the “natural” (flour is already refined) ingredients.

  3. PlasticPaddy says

    Publication in a peer-reviewed journal submits an article to external review, intended to expose errors and omissions. If the author wants appreciation (or kind and constructive criticism), she can submit the paper as a preprint to a select circle, either before or instead of submitting to a journal. If you mean that not enough balanced or appreciative assessments are published, I think there are Festschrifts and journals that publish review articles, where the accent is not on criticism and there is no incentive to downplay one person’s work in order to make another person’s (usually one’s own) work appear to be more original or significant.

  4. Well, I’m not an academic, so I don’t have an informed opinion. I’m just tired of agonism in general.

  5. I mean, I’m not saying people should only say nice things; I’ve frequently posted harshly about lousy writing and editing. But I didn’t start out looking for things to attack, I started out hoping to learn something; after extending the benefit of the doubt and being continually disappointed, I let fly. Case in point.

  6. [Comment deleted at author’s request.]

  7. PlasticPaddy says

    I am not sure there is a place for investigations of stereotypes in academic publications, except in investigating what level of belief (e.g., “strongly agree”, “no opinion”, “strongly disagree”, etc ) people assign to such beliefs or how stereotypes are propagated and reinforced. Even with the best intentions, investigations of stereotype “substance” will, I feel, fall foul of more or less unconscious bias or selection effects (as you seem to be saying).

  8. Tannen’s “‘ritualized adversativeness'” is characteristic of, at most, a particular kind of academic article. It is not, for example, standard in the sciences and engineering (which constitute by far the majority of academic publications by number, if not necessarily by word count). Nor is it the case in all the humanities. In history, for example, when someone is writing about previously-minimally-studied events, there is no previous narrative to rebut (although the adversarial tone Tannen describes does tend to emerge when historians write about better-known people and events, about which there is already a literature).

  9. I’m glad to hear it!

  10. David Eddyshaw says

    Not the case in medicine, either; though our journals might be improved by a bit more ritualised adversativeness, at that …

    “Though many previously published studies purport to deal with the question of the effectiveness of X, they are without exception the work of mountebanks and arrant charlatans unworthy of the sacred name of Physician …”

  11. I meant to include medical articles under the general heading of “science and engineering.” Whether that is the most sensible approach is questionable, but it seems pretty conventional. For example, around 2001, the once-vibrant news organ Newsweek ran a series, highlighting their picks for the ten best people in a number different areas. They lumped science and medicine together, and five or six of the people they named appeared to be primarily clinicians. Apart from the lack of usefulness of comparing heart surgeons biochemists, the choices were generally pretty dubious. There were (I think) only two honorees in the physical sciences: a prominent geologist and as astronomer whom I had never heard of. The magazine did at least acknowledge my complaint (although they didn’t actually print my letter, just summarizing it as the response of “a MIT physicist”) that they had completely overlooked the area of solid state physics,* which had such a huge—and growing—influence on our daily lives.

    In any case, there are some famously combative contributions to the medical literature. In one of Semmelweis’s open letters, he wrote:

    The murder must cease, and in order that the murder ceases, I will keep watch, and anyone who dares to propagate dangerous errors about childbed fever will find in me an eager adversary.

    * Solid state (or condensed matter) physics really has two major components. Physical Review Letters has sections for the major areas of physics research; however, until they reshuffled them a few years back (as part of a change aimed mainly at cutting back on the predominance of solid state physics articles in the journal), there were two whole condensed matter sections, one for structural properties and one for electromagnetic properties. Obviously, although the two are closely related, it is the latter category that is of central importance in the development of modern technology (and thus the one I was complaining that Newsweek had missed). This divide into two sub-subfields is also evident in other ways. There are two very popular graduate-level textbooks in condensed matter theory, Introduction to Solid State Physics by Kittel and Solid State Physics by Ashcroft and Mermin. Ultimately, the both cover the same material, but it is obvious that the writers had quite different ideas about what was the most fundamental aspect of the subject. Kittel** does everything you can possibly do with just the crystal lattice before introducing the electrons, while Ashcroft and Mermin do*** everything you can possibly do with just the electrons before introducing the crystal lattice.

    ** Like Jackson’s Classical Electrodynamics or Goldstein’s Classical Mechanics, these graduate texts are known primarily by the authors’ names, rather than their actual titles.

    *** Either “do” or “does” works here for me, depending on whether “Ashcroft and Mermin” is taken to refer to the authors, or metonymously to the book itself.

  12. [Comment deleted at author’s request.]

  13. Craig: it’s certainly a valid study. The sample might not be representative of every crowd in that demographic, but 2½ hours of recorded unprompted observation is substantial, especially for those days, when long recordings were rare, and conversation analysis was a new science.

  14. David Eddyshaw says

    I have a medical friend with an enviable ability to take a step back from everyday medical situations and notice something about them which is odd once you take a moment to reflect, and worthy of investigation – and publication.

    Not all worthwhile studies are based on huge series, and the plural of anecdote is data.*

    * I first came across this as “the plural of anecdote is not data.” However, this statement is erroneous.

  15. PlasticPaddy says

    I think a corollary is : the plural of selective/biased anecdote is selective/biased data or more succinctly, GIGO 😊. Some of my best friends are medics.

  16. tannen is doing a kind of analysis here that is based on close examination of a very thoroughly observed (and documented) set of interactions. that tends to be the kind of analysis that’s necessary to understand even phenomena that are amenable to large-scale statistical methods – there, it’s generally what has made possible a hypothesis to be tested that way.

    attacks on these kinds of analysis as “anecdotal” and thus to be automatically dismissed, whether aimed at linguistic analyses like tannen’s, focused observation of the behavior of a specific group of animals, or close textual reading, are one of the purest versions of the “ritualized adversativeness” that tannen examined*. and that’s just as true – and probably more so – when those attacks spill over from academic blowhards like alan sokol into the popular discourse.

    * generally deployed (quite consciously) in the u.s., at least, as part of the overall right-wing attack on comprehensive education, and the humanities in particular.

  17. BTW, from the very little I know about animal behavior studies, a lot of what’s published in that field is along the lines of “we saw X doing Y, which has never been observed before…”

  18. David Eddyshaw says

    It reminds me of the old chestnut about consultants (i.e. senior medical staff), frequently cited in olden time by their trainees:

    If a consultant has seen one case of a disease before, he says “In my experience …”
    If he has seen two cases, he says “In my series …”
    If he has seen three cases, he says “Time after time after time …”

    Actually, this is not utterly far from the mark.
    In my own line, there are so many rare diseases that you come across rare diseases all the time (just not the same one twice.)

    I once had occasion to ring up an internationally acclaimed expert in paediatric medical retina work about a case in a fifteen-year-old boy that was wholly unlike anything I had ever seen before. She said, after I had finished my description of the signs: “Oh, there’s a little epidemic of this just now.” I gathered that this meant she had seen two cases very like it in the past year. Still, that was two more than anybody else …

    (The child made a complete recovery.)

  19. John Cowan says

    The Harvard Law of Animal Behavior: “Under [well-]controlled experimental conditions of temperature, time, lighting, feeding, and training, the organism will behave as it damn well pleases.” A shorter version speaks of “well-controlled experimental conditions” without specifying them and of “a well-trained animal”.

    you come across rare diseases all the time

    “Rare conditions are extremely common.”

  20. PlasticPaddy says

    It is up to the publisher and reviewers to judge validity of any particular approach and extendability of the conclusions based on it. I have the impression that, in social sciences, experimental protocols used e.g., in medical science experiments (e.g., repeatability, blind or double-blind trials, control of sample to avoid bias, etc.) are not universally applied. When investigating the “truth” of stereotypes, I would expect extreme care, in order to eliminate bias. I suspect that the effort one might have to expend, in order to construct and apply a really robust experimental design, and the difficulty of interpreting the data in a demonstrably unbiased and non-selective way, might discourage researchers from doing, and journals from publishing more rigorous studies whose objective is to discover the truth behind stereotypes.

  21. J.W. Brewer says

    The penultimate paragraph of Tannen’s piece commendably raises the obvious points about the hazards and uncertainty involved in extrapolating from this small and non-random sample to larger population groups, with its sentences beginning “This raises the question of the extent to which …,” “My hypothesis is that …,” “My impression is that …,” “I suspect that …,” and “I have not encountered evidence to indicate that …” All of which nuances are followed by the final paragraph’s conclusion that More Research Is Needed.

    That said, perhaps holding all that stuff for the end (not to mention giving the piece an unhedged title) increases the chances of casual readers (and/or readers of summaries of the piece by third parties) simply assuming away all of those complications and treating this is a more definitive/essentialist finding about “New York Jewish Conversational Style” than it actually is.

  22. Well, that’s their problem. As PlasticPaddy suggests, more “rigorous” studies are unlikely to get done, and I’m not a fan of the “if there’s any way this could be misinterpreted or used for purposes of flimflammery, it shouldn’t be published” approach. I believe in the naked-lunch theory: let everyone see what is on the end of every fork.

  23. Stu Clayton says

    I believe in the naked-lunch theory: let everyone see what is on the end of every fork

    Whatever it is, it will vanish in a few seconds and be replaced by something else. You will be none the wiser.

    I am a proponent of boxed lunch theory. Let everyone take out and share only what they want to. That way people who can’t afford lunch get something to eat, and those who can get to boast.

  24. J.W. Brewer says

    I think anecdotal stuff like this (in this particular field as of that time period, at least) is worth publishing precisely because it may serve as a prolegomena to and encouragement for more rigorous studies of the same phenomenon, with larger and statistically valid datasets. But if that never happens, the anecdotal results do not suddenly become meaningful on their own, and remain in that sort of liminal interesting/suggestive zone where they could ultimately be consistent with lots of different true generalizations about the world, not all of which are consistent with each other.

    You could take the exact same conversation and describe it as suggestive of the (not empirically implausible) theory that echt-Californians of the Seventies (with all their stereotype-confirming weird-beard “mellow” folkways) are the marked group here, with idiosyncratic expectations about how conversations ought to work when compared to the default/unmarked expectations of normal-American interlocutors.

  25. Stu Clayton says

    lots of different true generalizations about the world, not all of which are consistent with each other.

    That is a strange idea. Do you have an example to hand of two true statements about the world that are not consistent with each other ? I’m not sure where consistency comes in here, or inconsistency either for that matter. As for truth, I’ll let it pass for the sake of argument.

  26. J. W. Brewer says

    Not so much a strange idea as a regrettable editing glitch not noticed in time to cure. Either delete “true” or add •potentially” before it.

  27. jack morava says

    @ JWB,

    I believe many assertions about truth are about what’s generically the case, and it can certainly happen that assertions A and B are both generically true, but cases in which both are true at once are relatively few.

    A guppy is neither a good generic example of a fish nor of a pet, but it is a good example of a pet fish.

  28. @PP: yes, that’s exactly what researchers in many fields (including linguistics) have spent over a century (more like five to fifteen, depending on your frame of reference, really) doing!

    and, perhaps a bit contra our delghtful host (to play against type for a moment), the fantasy that other methodologies are somehow more “rigorous”, is what i might start calling Tannen’s Agonism.

Speak Your Mind