Dear Sarah,
I have been having a look at your data and trying to determine exactly where the problem lies and what your expectations are (and at this point, I realize, maybe it will suit the other members better if we talk about this over our own email addresses, since it may be more a data issue than an issue of principle).
The question is what your expectations are, based on your auditory analysis. Here's my attempt at guessing: Assuming O refers to older speakers and Y to younger, what you want to be able to show with the visualization here is that these tokens of PRICE (most clearly with the young 'green' speaker) are more front than TRAP and have a height variation such that most tokens of PRICE are more close than TRAP, while for the older speaker PRICE could indeed be more open than TRAP. Assuming the blue speaker is also a younger speaker, though, it's not clear that these two younger speakers (blue and green) behave similarly, so the question is also back to you. What are the expectations for what the visualization should show?
One other issue is which version of W&F you are using, since m(W&F) (also available on NORM) does slightly different things. But we can get back to that, off-list if you (or others) would prefer...
Looking forward to hearing more.
Anne
-----Original Message-----
From: Variationist List [mailto:[log in to unmask]] On Behalf Of VAR-L automatic digest system
Sent: 5. maj 2012 01:05
To: [log in to unmask]
Subject: VAR-L Digest - 1 May 2012 to 4 May 2012 (#2012-65)
There is 1 message totaling 97 lines in this issue.
Topics of the day:
1. Normalisation of the PRICE vowel
----------------------------------------------------------------------
Date: Fri, 4 May 2012 15:48:18 +0100
From: Sarah Haigh <[log in to unmask]>
Subject: Normalisation of the PRICE vowel
Hi all,
About a year ago I was struggling with some vowel normalisation woes, and received a lot of very helpful advice from many members here, for which I was very grateful. However, after further discussion and experimentation (and a health-related absence from study), I'm back with further questions!
This case in particular relates to a monophthongal pronunciation of PRICE, normalised using the Watt & Fabricius method through NORM suite. I have various outlying values which seem to affect the normalised result - if I remove these values, the results change to something that, in fact, looks more plausible. Examples can be seen here:
Outlying results included: http://oi45.tinypic.com/2rf5ws7.jpg
Outlying results removed: http://oi49.tinypic.com/2m84cav.jpg
This has happened with many of my results for this vowel, both monophthongal and diphthongal. Obviously I don't want to just get rid of these outliers and ignore them, but how can I trust my results if they can be changed like this? Am I doing something wrong, or do I need to use a different method? As you can see, there's a lot of overlap between PRICE and TRAP, which I have been concerned might be causing problems. Is there anything I can do, or do I just have to abandon (again) my attempts to get any normalised results for this vowel?
Many thanks for your time, and any advice you can offer!
Sarah
########################################################################
The Variationist List - discussion of everything related to variationist sociolinguistics.
To send messages to the VAR-L list (subscribers only), write to:
[log in to unmask]
To unsubscribe from the VAR-L list, click the following link:
http://jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=VAR-L&A=1
------------------------------
End of VAR-L Digest - 1 May 2012 to 4 May 2012 (#2012-65)
*********************************************************
########################################################################
The Variationist List - discussion of everything related to variationist sociolinguistics.
To send messages to the VAR-L list (subscribers only), write to:
[log in to unmask]
To unsubscribe from the VAR-L list, click the following link:
http://jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=VAR-L&A=1
|