[PLing] Erinnerung: Vortrag Yanovich (8. 6.), "Evolutionary methods in historical linguistics: forward-time and backward-time analysis"

[Laura Grestenberger]

Liebe PLing-Liste,

Ich möchte gerne nochmals an den morgigen WSG-Vortrag von Igor Yanovich 
(Tübingen, PI DFG Emmy Noether research group "Modal systems in the 
historical Slavic languages", 
http://www.sfs.uni-tuebingen.de/~yanovich/) erinnern. Der Titel ist 
"Evolutionary methods in historical linguistics: forward-time and 
backward-time analysis" vor. Das Abstract finden Sie in der Ankündigung 
im Anhang und weiter unten in der Email.

Der Link zum Raum:
https://univienna.zoom.us/j/91578136358?pwd=b1FmUXJjeVQ3RnVWVk5YUytHdlEwdz09

Meeting-ID: 915 7813 6358

Kenncode: 564019

Wann: 8. 6., 18.30h

Interessierte sind herzlich willkommen!


Mit freundlichen Grüßen,


Laura Grestenberger


----------------------------------------

"Evolutionary methods in historical linguistics: forward-time and 
backward-time analysis"

Evolutionary methods in linguistics have reached the stage where they 
have a dedicated regular conference (EvoLang) and, arguably, at least 
two journals (Language Dynamics and
Change and Journal of Language Evolution). This subarea has developed 
from pioneering early monographs breaking the way (to give just two 
examples, Croft 2000 and Niyogi
2006) to the multitude of papers published every year by an ever growing 
community of scholars. However, we still have to reach the next level – 
that of a solid synthesis that can
help the field to consolidate the gains and to better focus future work. 
In this talk, I try to make a modest contribution to this synthesis by 
demonstrating how two big analytical
frameworks of mathematical population genetics can be fruitfully used to 
think about a wide range of linguistic problems. My focus will be on how 
a wide range of linguistic settings can, sometimes surprisingly, be 
explored using rather uniform mathematical theory.
Crucially, this would require little new fundamental research: only 
sensible application of already reasonably well-understood math to 
linguistic problems. The first framework we will look at is 
»forward-time analysis«. Starting from a certain set of objects that can 
be inherited through time, the relevant mathematical theory studies the 
trajectories that this set – this »population« – can take in its future 
developments. The laws that govern such development crucially depend on 
the properties of the specific inheritance process at hand, and there is 
great variety of such processes both for biological organisms and for 
different linguistic objects in language change. Knowing this is useful 
when thinking about whether, and which, evolutionary »machinery« suits a 
particular languagechange problem. To illustrate this, I introduce a 
vignette on how the predominantly unidirectional nature of 
grammaticalization can be derived from the first principles for a 
clearly defined, but wide range of linguistic settings.
The second line of analysis is »backward-time analysis«. It is a younger 
subfield within population genetics, starting crystallization in its 
modern form just in the early 1980s, under the label of »coalescent 
theory«. This branch of mathematical population genetics starts from a 
present-day sample of objects that were subject to inheritance, and then 
asks which genealogical relationships these objects could have had with 
each other in the past. Somewhat surprisingly, this backward-looking 
perspective obtains relatively easily some very general mathematical 
results. Moreover, it allows creating computationally tractable
methods of studying very complex models. I illustrate how this can be of 
use to historical linguists by presenting the results on the Bantu 
language family’s prehistory that we obtained jointly with Silvia 
Ghirotto (U Ferrara), Patricia Santos (U Bordeaux) and Andrea Benazzo (U 
Ferrara).



-- 
Mag. Laura Grestenberger, Ph.D.
Institut für Sprachwissenschaft
Universität Wien
Sensengasse 3a
1090 Wien
+43 1 4277 41720
-------------- n�chster Teil --------------
Ein Dateianhang mit Bin�rdaten wurde abgetrennt...
Dateiname   : WSG_Yanovich_210608.pdf
Dateityp    : application/pdf
Dateigr��e  : 128710 bytes
Beschreibung: nicht verf�gbar
URL         : <https://lists.univie.ac.at/pipermail/pling/attachments/20210607/a0787de0/attachment.pdf>