Language Change – Palatization Survey

As we begin to look at the topic of Language Change for ENGB3 and start thinking about research topics and methods for ENGB4, my Y12 English Language students are looking at changing patterns in the palatization of alveolar consonants followed by the palatal continuant [j]. In many words the pairs of phonemes are being replaced by a single affricate.

The survey sheets they are using are here. They are (deliberately) far from perfect. One of the things we’ll be considering next week is how effective this is as a research methodology, and what changes we might make to produce better results.
This entry was posted in Uncategorized and tagged , , , , , . Bookmark the permalink.

13 Responses to Language Change – Palatization Survey

  1. Sarah Kenny says:

    The methodology was quite flawed. If we asked people to repeat the words after us they might have changed their pronunciation to make it closer to ours, this would give inaccurate data. If we showed the participants the words on the sheet and then asked them to say them, they would have been able to see the columns and the tally charts. This would mean they could try and gues the aim of our survey and respond to any presumed demand characteristics- again giving inaccurate, unreliable data. We could have written the words out on a seperate sheet and then shown them to the person, but even then if we ever asked them to repeat their pronunciation for our sake, they might think they were saying it ‘wrong’ and change their pronunciation. We should have written all the words out on a seperate sheet and then recorded all the individuals pronouncing them so that we could listen to it as many times as need be.

    Investigator bias was also incurred, as random sampling was not used. Random sampling would give more generalisable data.

  2. Emma Gyarmati says:

    I agree with Sarah, in her views of how the methodolgy was flawed. It also had its problems as some of the words that were on the sheet were words that some of those that we asked had perhaps no heard of, so didn’t even pronounce in the way we were looking to hear, so did not give a true reflection of the investigation. Also some of the words chosen were not appropriate for those we were asking for example; if using a grandparent the word bisexual may not be comfortable to use, so you may skip this one out therefore not having reliable data. We could have used more common words and familiar words so that all the people we asked were more familiar and confident to pronounce the words, because if you see a word for the first time they may feel pressured to answer in the way we are looking for so not giving a true reflection.

  3. Amy Kilner says:

    I also think that the methodology could be flawed as human error could be a factor which influences the results of the investigation. The individual who is carrying out the investigation will have to determine which pronunciation the person has used. This is quite difficult as i find it difficult to distinguish between the three choices when the full word is spoken aloud. The person who is being surveyed may have to repeat the word several times to make the data more reliable or as sarah suggested the person could be recorded saying the words so the sample could be listened to as many times as required.

    Im not sure whether this is the right idea, but the words which were chosen for the survery are words which are not normally used in everyday conversation. This may be because the pronunication of them will be individual to the particular person carrying out the survery and may be less likley to be influenced by other factors which have changed common, every day words. This would make the investigation more reliable as the pronunciations will be more individualised to the candidate.

  4. Mr Heald says:

    Impressively thoughtful responses from all three of you so far. I think I need to get you to teach me research methodology, Sarah!

    There are so many different ways of conducting research (and of getting it wrong!) that there’s still plenty for the rest of you to say. I look forward to your comments…

  5. Eilish says:

    I agree with Sarah, when she talks about how the person taking part in the investigation may copy the way you say the word, and therefore the data becomes unreliable. I found this with two people in my investigation. When one of them didn’t know how to pronounce the word they would ask me how to pronounce it, and would copy exactly how I said it. Also another person became very self concious about how she was pronouncing the words compared to the person sat next to her, and therefore copied the pronunciation of the person sat next to her. This made the results of this particular candidate irrelivant to the investigation.

    One way to solve this would be to give the sheet with the words to the candidate and let them say the words only once, and maybe record them saying it. As during the investigation, when I asked someone to repeat the word, they became self concious and pronounced the word differently the second or third time. I would also make them do it alone, so they would not be infulenced by those around them.

  6. Sam Dunstan says:

    Since I sent this to Mr Heald instead of posting it here i figured i should go ahead and post so here it is:

    After looking at the data from the survey you gave the class last lesson I’ve come up with another idea as to how you could test the pronounciation (Yes Mr Heald I realised my mistake, a muggled up what the aim of the survey was with what we were doing with Miss Hampshire thankyou for drawing attention to my error) of those words. Why not get participants to be in a semi-structured recorded interview. By semi-structured I mean there is a specific topic of conversation and a few questions that will lead the subjects to say those words. And by recording the interview and listening back you can record on the survey sheet what type of pallet was used for each word and even play the recording for others to check that what ou perceive to be what was said is correct. The added bonus of this being with a recording you can rewind and double or even triple check your results!
    And to make sure all the selected words were mentioned the interviewer can tick off the words on a checklist whilst the interview is taking place and if a certain word is not mentioned they can ease the participant into saying it :).

    For this method to work you would need to use some level of deception as we talked about last lesson and similar to Eilish’s problem which we spoke about in lesson and not tell the participants the true aim of the interview until after the experiment is conducted and the data has been collected. This would be because the participants would change the way in which they pronounce the words if they knew you were investigating this so by lying and saying you are investigating something different they would use there own standard pronounciation. This might raise some ethical issues by decieving participants but I think the validity of the results out weighs the ethics of the experiement and providing you remember to explain the true aim of the investigation after the interview has taken place im sure there would be no issue with it.

    Any other strengths or problems people can see with this??

  7. Zerin Pekin says:

    Sorry for the late reply.
    I think everyone has already mentioned any flaws that I could think of. Similiarly to Eilish, whenever someone asked me how to pronouce something, they would copy the way I said it. In lesson today we decided that this problem could be overcome by pronouncing the word with each different phoneme in it and then let the participant repeat whichever they felt most comfortable saying. At first, I thought that this was quite a good idea but after thinking about it for a while, I thought this might bring about demand characteristics (as mentioned by Sarah) because the individual may feel self-conscious about the way in which they’re pronouncing the word and worry about saying it ‘wrong’. In fact, I believe the entire survey as a whole could cause demand characteristics to occur. This is because, if you ask someone to pronounce a particular word, they automatically feel self-conscious about the way in which they’re pronouncing it. For this reason, I liked Sam’s idea of using a semi-structured interview and recording it. The results would be a lot more reliable because the participant would be unaware of the true aim, which would reduce demand characteristics and also, a discussion is more natural than a survey and so their would be higher ecological validity to the results.

  8. Doug Belshaw says:

    I’m going to throw the Hawthorne effect into the mix here as well. Always something to be wary of!

    Say hello to Mr Belleini for me. 🙂

  9. Kevin Watson says:

    In phonology this process is sometimes called ‘yod coalescence’ – ‘yod’ is the name given to the /j/ sound, and ‘coalescence’ means to ‘merge’. There’s another process called ‘jod dropping’, which happens in some accents of English but isn’t as widespread as coalescence. It would need to go in the ‘other’ column of your questionnaire. You can read about yod dropping here:

  10. Kyle Thomas says:

    I was beaten to mentioning the hawthorne effect, something I noticed on my investigation. When my participants were made aware of their pronunciation as I was asking them, I felt that they pronounced the words differently to how they would have usually pronounced it. Their pronunciation changed because I was observing it so the study lacked validity. The study (I’m just pulling things from psychology here) lacked ecological validity as well as we’re not asked to pronounce words in real life hence the results are not representative of real life pronunciation. To combat both these issues perhaps just engaging someone in conversation without disclosing the aim and hoping they say the words on your sheet would be a better method.

    Another thing I noticed was how big a variable the investigator was in the study. My interpretation of what phonemes the participant used may be different to what someone else would think. If the study was conducted by someone from the south testing the same people I used, I think we would have got different results due to the different phonemes our ears would possibly be more acute to distinguishing.

  11. Helen says:

    Everyone’s definitely pilfering stuff from psychology now which was all I could think to write about.. I like Sam’s idea about the semi-structured interview although I think some of the words would be really hard to get into a relatively natural conversation.. like ambrosia.. Also, you’d assume that the interview would be fairly time consuming, and something was said in class about having to conduct the survey on as many people as possible for it to be reliable in any kind of way, so overall I think it’d take too much time, but the idea of recording it was really good so you can let other people listen to it to validate any results.

  12. Ryan Bailey says:

    Like most of the other comments on here, I think the main problem here is the fact that some words may be difficult to pronounce, and if the participant can’t pronounce it, the word cannot be said to them as this would compromise the whole investigation. I think another possible problem is the fact that people may change their pronunciation based on the fact that they are being tested (ecological validity for all you who do psychology) and this may be another big problem, as even if they are not told it is an experiment the fact that they are just reading out words from a page may be a bit suspicious to them. The only way I can think to get around this problem is to listen out for words on the list whrn they are talking in everyday conversation, but this poses the obvious problem that hardly any words (if any) are likely to be said in a conversation. Also, since the sounds are so similar, when the person is talking quickly, it is very easy to miss a sound, and you can only say ‘can you repeat that please?’ a few times before people start to get annoyed/suspicious!

  13. Daniel Marsden says:

    After many attempts to post, here it is.

    After struggling to think of a comment to post without wanting to repeat anybody else, I have come up with an idea our survey has not taken into account the area of where the participant is from.
    Would the pronounciation of these words be any different if the participants were from the south?
    If so, our surveys results would not generalise to the population and maybe if we used participants from varying areas then we could link different pronounciation to the area which a person is from.
    Just a thought.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s