“Conversational style” in literary works
Yesterday I stopped with this quote from Joe Peschio:
The argument runs that an illusion of conversation between author and reader is maintained by borrowing certain conventions and features of real spoken discourse and representing them textually. The main obstacle here is laughably obvious: lacking recordings, we have no idea how people talked two centuries ago, and a priori definitions of conversational style are therefore an analytical dead end. (47)
I’m still thinking my way through this, and I should read a lot more about it, but on the whole I don’t think it’s true that “we have no idea how people talked” in the early 1800s. We don’t know everything about how they talked then, but then we don’t know everything about how everyone talks now.
Imagine a rural Russian had invented a wax cylinder recording system in 1800 and recorded lots of conversations, but then a natural disaster destroyed the whole village, kept word of the invention from getting out, and preserved the recordings. Tomorrow someone finds them. How much would it change our opinion about whether something written in 1800 mimicked actual conversation of the period?
We might learn some interesting particulars — I’d want to listen — but even something as concrete and “objective” as a recording wouldn’t solve all our problems. It would always be possible that Linguistic Element X from literary conversational style was used in those days in speech in at least one corner of the culture, but didn’t happen to be captured and preserved; the creator of the recordings might have selectively destroyed the interestingly non-standard ones; and people may have changed they way they talked because they were being recorded. We have movies, radio, TV, and documentary recordings of 1950s English speech, and this is useful, but we have to be careful how we interpret individual examples of recorded 1950s speech, much as we do with transcribed 1950s speech.
Also, readers have strong intuitions about whether written words sound like they could have been said aloud, and this is true even when they’re reading the purported speech of people who lived before their grandparents were born. It may be difficult, as a practical matter, to test these intuitions against nonexistent recordings of nineteenth-century conversations, but I suspect they’re intersubjectively verifiable — we’re mostly going to pick the same authors as the ones who seem to be transcribing real speech they heard, or as the ones who seem to be making stuff up about a linguistic milieu they’ve never been in.
Isn’t Boris Bukhshtab probably correct when he says that Dal’ and Vel’tman made a literary representation of the speech of the common people that was “actually based on popular dialects,” while Masal’skii’s commoners spoke in a conventional and inauthentic “literary Esperanto”? None of us, Bukhshtab included, were there to hear how peasants talked in Vel’tman’s day, but there are several lines of indirect evidence we can use to substitute for the missing recordings of actual speech. We can use the spoken and written Russian (and perhaps Ukrainian, or other languages of the relevant part of the Empire) of later periods to reconstruct the spoken Russian of the early nineteenth century, the way people use taboo words in Romance languages to recreate unwritten classical Roman swearing. We can check how contemporaries who did know how (some) people talked in those days reacted to different examples of “conversational” writing. We have contemporary metalinguistic laments about the difference between actual speech and written convention. We can look for text-internal evidence, such as whether non-standard phonological, morphological, or syntactic features occur in a regular way, or we can compare a text to contemporary literary simulations of conversation to see if it is idiosyncratic or resembles writing we consider authentic or writing we consider conventionalized.
I’d even say that learning any language involves repeating that last process over and over: we hear a phrase that stands out from our favorite writer, or a stranger at a bus stop, or a younger cousin, or a docent at an art museum, and in our minds the sociolinguistic and stylistic coloring of the phrase takes on a little of our opinion of the speaker and setting, while our opinion of the speaker and setting are slightly recalibrated depending on our pre-existing idea of the phrase. I assume Bukhshtab believed that ежели, таперича, вестимо, чай, and право-слово were conventionalized markers of peasant speech that traveled from one literary work to another because he had seen them in a whole network of texts, some of which he considered inauthentic for other reasons.
No two generations or even people will hear/read the same words and have the same attitude toward them, which explains semantic change over time and idiolectal differences. But communicative situations overlap with each other. Without even trying, if we’re exposed to enough different kinds of language, over time we develop a sense of how Linguistic Element X was used and perceived by people of different ages and social groups in 1810 writing, 2015 speech, and 1965 radio broadcasts. This lets us guess how it might be used and perceived in a linguistic environment we have no direct access to, like 1810 speech, or for that matter like the majority of 2015 speech that will not be preserved and that we will never encounter.
Long story short, I’m pretty sure that recordings of 1810s speech would help less than Peschio implies in that one sentence, but that the written sources we do have let us infer more than he acknowledges about the spoken language we don’t have.
See Joe Peschio, The Poetics of Impudence and Intimacy in the Age of Pushkin (Madison: University of Wisconsin Press, 2012).