"dog whistles"

[Geoffrey Nunberg]

If you were going to botanize this device according to traditional rhetorical categories, metonymy seems to me less apt than synecdoche, in its less common meaning of "whole for part.” To qualify as a dog-whistle, a statement has to be ambiguous between a literal meaning and a potential conveyed meaning that the speaker can plausibly deny. That means that both meanings have to be of the same type-- e,g. "hoodie wearers," "food stamp recipients" or "the bling crowd" (once used by National Review) for blacks; "city boy” for Jews, etc. where the denotation of the conveyed meaning is stereotypically a (proper) subset of the denotation of the literal meaning, so that the speaker can disavow the implicature, e.g., "not all hoodie wearers (inner-city inhabitants, etc.) are black.”

So these could be described in classical terms as "potential whole-for-part synecdoches.” But as with euphemism, etc., the speaker’s (real or imputed) motivation is an independently crucial criterion. That is, the figure is designed to avoid unambiguously suggesting certain social attitudes to listeners who disapprove of them (as distinct from euphemisns, which enable the speaker to avoid uttering a coextensive term that some listeners find unsavory).  “Obliquity” conveys one part of this, and “conivinutation” nicely conveys the other, though neither is a word they would let you use on public radio.


Geoff

>
>
> From: "Joel S. Berson" <Berson at ATT.NET>
> Subject: Re: "dog whistles"
> Date: April 4, 2014 at 6:44:27 AM PDT
>
>
> At 4/4/2014 07:42 AM, Amy West wrote:
>> On 4/4/14, 12:00 AM, Automatic digest processor wrote:
>>> I like the description the philosopher John Holbo at Crooked Timber
>>> has used, "impolite fictions," but that doesn't get at the semantic
>>> process here, which it seems to me to involve referring to X via
>>> one of its stereotypical properties (as, e.g, "inner city," "food
>>> stamp users") with the intention of evoking but not actually
>>> denoting it. (Or maybe I should make that, "referring to X by
>>> naming something to which X stereotypcially applies -- e.g., food
>>> stamp users are stereotypically black.) But what should it be called?
>>>
>>> Geoff
>> It seems like a "stereotyped metonymy" to me, where a stereotyped
>> quality is used to stand for the whole stereotype. Or a "false metonymy"
>> to be more explicit.
>>
>> ---Amy West
>
> Isn't metonymy already a stereotype, by choosing one characteristic
> to be a "type" (as they used to say) of the whole?  Some adjective
> needs to be added that expresses the "impolite" (derogatory, hostile)
> aspect of the "dog whistles".
>
> Joel

------------------------------------------------------------
The American Dialect Society - http://www.americandialect.org