Sunday 16 February 2014

Work | An example of what I am looking for in my project

I have just finished working through some 5-6,000 historical letters to/from Irish emigrants. It took me a good few months and it feels great being done. Now remains analysing the results. And writing. I need to continue writing. I know. And I am. In the meantime, I thought I'd share with you some examples of what a letter might look like and what kind of features I'm after.


Letter example

The letters in the Corpus of Irish English Correspondence (CORIECOR) have all been digitised and appear for instance like this:



New York March 28th 1848

Dear parents I am going to wright [write?] those few Lines
to ye hopeing [hoping?] to find ye In as good State of health
as this Laves [leaves?] me in at Present thanks be to God for
his mercy to me Dear Parents I am sending ye two pounds at
Present but I had no opportinity [opportunity?] of sending ye
no more but yet I will not forget for to send ye some more
assitance [assistence] as soon as I can Dear parents let not
ye be uneasy abot [about?] me for I am in good halth [health?]
thanks be to God for it Dear father and mother I was verry
[very?] uneasy Abot [about?] ye since I left home knowing the
state ye were in And how yer [your?] Sircumstance
[circumstance?] was
[...] (Eliza Quin, 28.03.1848)

You'll notice a few things straight away: There is no punctuation, the capitalisation is haphazard and there is a lot of spelling variation (given that many of the writers lacked the necessary schooling/training we'll call it spelling variation instead of spelling errors). There are certainly many letters with 'correct' punctuation, capitalisation, spelling etc., but for my purposes, the more 'oral' or vernacular the letter, the better. 

As you can see, spelling variation is followed by a suggested interpretation in squared brackets. Could I have used a script to automatically search for these in the corpus? Well, yes and no. It's doable, but the corpus exists as yet as collection of text files in folders (and hasn't been made into a searchable database yet), and sometimes an instance of spelling variation has been missed by the transcribers.


Phonetic representation example

My interest here is spelling variation that reveals something about how the writers spoke, in other words: the historical phonology of Irish English. Here is a cut-out from a research poster I made for the PARLAY conference in York last year with some features of Irish English that are revealed in a letter identified and explained:


(Typo in 'Schwa-epenthesis: persusent')

The intention behind gathering this kind of data is so that current knowledge of the development of Irish English phonology may be added to. Well, I'll get back to that.

Also, this is intended as a light-hearted blog (I'm sure you can tell from the combination PhD project in linguistics / graphic design), so I won't go into too many technicalities. There are for instance some great quotes that I have collected from some these letters, just waiting to be shared with you...

No comments:

Post a Comment