Monday 27 May 2013

‘What did you see? I didn’t quite hair you...’

                               even toddlers can understand a foreign accent 

We all know how difficult it can be to understand somebody who is speaking in a different accent to our own.  This is hard enough as an adult at times, but imagine what it must be like for a child who is just in the process of learning language and pronunciation.  Rachel Schmale, Alejandrina Cristia and Amanda Seidl set out to investigate whether unfamiliar accents completely impede young children’s word recognition.  They were working on the premise that being exposed to an unfamiliar accent, even for as short a time as a minute, makes it easier to understand.  The idea behind this is that the listener picks up on patterns in the accent and then ‘maps’ them onto what they hear.  One previous test showed toddlers a picture of a dog whilst they heard ⁄dæg/ (dag) rather than the expected pronunciation and, similarly, a picture of a ball whilst they heard ⁄bæl ⁄ (bal). The toddlers were then found to ‘generalise’ this sound change when it came to other objects; so, for example, they looked at a sock when hearing /sæk/ (sak) but, interestingly, not when hearing /sɪk/ (sik) – showing that they had learnt that specific pattern. 

However, Schmale, Cristis and Seidl were aware that, in real life, children are not going to encounter a speaker with a different accent which only has one specific feature that is unfamiliar; they are much more likely to hear a foreign accent with many different features from their own.  The researchers wanted to face toddlers with exposure to a natural accent in the context of fluent speech and to do this they tested monolingual 2 year olds’ ability to recognise a newly learned word when it was spoken in a foreign accent.  Firstly some of the children listened to a passage of text, either in their own American English accent or in a Spanish accent.  After that they were tested with names of objects in the foreign Spanish accent to see if they could identify them.  The researchers tested this by tracking which objects the children’s eyes looked towards when they heard it being referred to*. It was found that the children who had been briefly exposed to the foreign accent beforehand were much more successful in this word learning task whereas the children who had not been exposed to the foreign accent were unable to do it. This suggests that even a few minutes of exposure to a different accent is sufficient to ‘tune’ our ears and help us understand it.

The researchers propose that we adopt two strategies when faced with an unfamiliar accent.  Firstly, we will shift our sound boundaries to accommodate to another’s accent – just as the children did in the example of dog and dag above. We will try to take a pattern and impose it onto other words to help us to understand them.  However, unfortunately, language is more complicated than this!  So, secondly, they propose that when listeners are faced with an accent that seems to differ dramatically from their own, the linguistic brain will relax its rules about pronunciation and accept a certain degree of deviation from its norms.   They admit that this could lead to misunderstandings as listeners will not only accept dag for ‘dog’ but may also start to accept things like beg for ‘peg’ and sit for ‘seat’, leading to lots of confusion!  They also propose the idea that a speaker’s ability to adapt to a new unfamiliar accent may literally improve with age – as we get older our vocabulary expands and, therefore, we have more resources to draw on when faced with a new accent.  What is certain is that this is a fascinating area that needs further investigation.

* This is easier to understand if you view the Youtube video at
Schmale, Rachel, Cristia, Alejandrina and Seidl, Amanda (2012) Toddlers recognize words in an unfamiliar accent after brief exposure. Developmental Science 15 (6): 732-738.

doi. 10.1111/j.1467-7687.2012.01175.x

This summary was written by Gemma Stoyle

Monday 20 May 2013

him/her, he/she, Ms/Miss…What do we use?

At times, the wealth of terms we have at our disposal to refer to someone can become confusing.  For example, should we be saying chairman, chairperson, chairwoman or just chair?  Which is correct and which might be offensive?  Most of us would hope that, over time, language use has become less sexist and one way of investigating this is through corpus based research.  This involves the analysis of large collections of computerised texts, identifying frequent linguistic patterns.  Paul Baker decided to compare four of these large collections or ‘corpora’ from 1931, 1961, 1991 and 2006 to explore whether male and female pronouns and nouns showed any signs of bias in language.
Baker found that, since 1961, there has been a decrease in usage of all male pronouns, especially  he, whilst female pronouns, such as she and her, have increased slightly.  Interestingly, the pronouns I and you have also increased in usage.  This may be due to the fact that written English is becoming more conversational and personalised over time, which could relate to the decline of third person pronouns in favour of the first and second person.

Another type of pronoun which Baker considered was that which attempts to include both males and females, for example him/her, he/she, he or she or s/he. The results showed that they are rarely used and although there was an increase in their usage between 1961 and 1991, the total for 2006 is less than half that of 1991. This suggests that they are becoming unpopular and in time, may even die out.  One reason for this could be that people find some gender neutral terms like s/he distracting or messy.  They seem easier to implement in writing rather than speech, which could prove to be a barrier to their long-term uptake.

When Baker concentrated on the nouns man, men, woman and women, he found that the four words were actually converging and being used as frequently as each other.  However, this wasn’t true when they were considered as affixes (i.e. as part of another word).  For example, the word spokesman still appears to be the most frequently used term, despite other equivalents like spokeswoman and spokesperson existing and the latter being used slightly more frequently in 2006 than ever before.  On analysing his data further, Baker found that spokesman is rarely, if ever, used to refer to a woman and he surmises that perhaps its frequency over other forms reveals the social reality that this is a role that men tend to carry out more than women.

Interestingly, when he analysed the frequencies of the similar type of word chairman / chairperson / chairwoman / chairlady or just chair, he found that, although chairman has always been and remains the most popular choice, there was an increase in the gender-neutral chair in the 2006 data, giving rise to the hope that it may start to replace chairman.  Compared to spokesperson, its popularity could lie in the fact that it is such a neutral term.  Spokes already exists as a completely unrelated plural noun and any word ending in –person can face resistance as it sounds so ‘politically correct’, which users often find off-putting.

Finally, Baker considered the titles Mrs, Miss, Ms and Mr, which have long been of interest to linguists in English-speaking countries due to their inbuilt inequality as labels.  Males are not forced to reveal their marital status with Mr, whereas females have to when they choose between Mrs or Miss.  Ms was conceived in the mid-twentieth century as an equivalent to Mr.  Nevertheless, apart from the confusion surrounding how it is pronounced, it is often connected with being divorced or a lesbian, thereby losing its neutrality.  Baker found that Ms was still very rarely used as a title but, perhaps more interestingly, that all the titles for both genders had decreased over time, so much so that he concludes that if the trend continues, all gender marked titles in English could become very rare in thirty years’ time.  In addition to people becoming more aware of gender inequality in language and the fact that fewer people are married now than in 1931, Baker explains a possible reason for this as being the increasing personalisation of British culture.  Therefore, rather than Mr Smith we may use William Smith or even Bill Smith instead, a much more personal and emotionally involved address. 

Baker concludes, reassuringly, that people are becoming more easily persuaded to stop using a sexist or biased term such as Miss.  He found that if a new term needed to be used then one that sounds more natural and is based on an existing word, such as chair, is more likely to be successful.  However, the invention of a completely new term, such as Ms or -person, is likely to be met with suspicion and resistance.

So in answer to the question of the title, it seems that maybe we’re beginning to NOT use any of the terms.  Such reassuring news is also supported by the fact that the terms feminism and feminist, which occurred just 3 times in the 1931 corpus, were found 59 times in 2006.  Good on her/it/them/us!
Baker, Paul (2010) Will Ms ever be as frequent as Mr? A corpus-based comparison of gendered terms across four diachronic corpora of British English. Gender and Language 4 (1): 125-149.

doi : 10.1558/genl.v4i1.125

This summary was written by Gemma Stoyle

Monday 13 May 2013

What’s new, pussycat?

 such a byootimis littlol kitteh!

 These few words are in Lolspeak, the language created by users of the website (formerly www.icanhas The language dates from 2005, when users of the website 4chan began to post lolcats every Saturday, renaming the day as Caturday. Lolcats are pictures of cats with captions written in playful, idiosyncratic English, now known as Lolspeak or Kitty pidgin. The cheezland website is now part of the Cheezburger Network, thought to have a fan base of at least 24 million people. There is now even a Lolcat Bible.

Why do so many people find this way of communicating attractive? The users, who call themselves cheezpeeps, claim that it is simply ‘silly fun’. Ilaria Fiorentini, though, suggests that it is more than this: through Lolspeak, she says, people playfully manipulate their language to construct an online fantasy group community and an identity that is simultaneously both of a cat and themselves.

Fiorentini analysed 1067 comments (17, 195 words) on the icanhazcheezburger website. She argues that although there are Lolspeak glossaries to help people use and understand Lolspeak, many of the features depend on the creativity of its users.  Interestingly, the elaborate language that cheezpeeps are constructing is evolving through processes that are typical of ‘normal’ language change.

One such process involves regularizing English verb forms. Whereas standard English has a present tense –s suffix with third person subjects only, Lolspeak – like many English dialects – has the –s suffix with all subjects of the verb (e.g. we awl needs sumfing tu gib us teh comfort, ‘we all need something to give us comfort’). In Lolspeak the regularization process is generalised still further, so that the suffix also occurs with past tenses (ai jus hadz a baff, ‘I just had a bath’) modal verbs (yu cants be a nan teak, ‘you can’t be an antique’) and infinitives (awl ob dose, and mebbe sum ivys oar fernz tu puts aroun it, ‘all of those and maybe some ivies or ferns to put around it’.

Cheezpeeps also make irregular past tense verb forms more regular: examples are kommed in wi kommed home, ‘we came home’, and seed in I nebber seed a kitteh do wat dey otter, ‘I never saw a kitten do what they ought to’. Double marked past tenses also occur in the posts that Fiorentini analysed, like wented in hubcat and ai wented tu the grossree store today, ‘my husband and I went to the grocery store today’. As she points out, regularisation of this kind is a very frequent phenomenon, typically occurring when children acquire a first language and when we learn a second language.

‘Lexicalisation’ is another typical process of language change, where a phrase becomes used as a fixed word: a well-known example is the word goodbye, from the phrase God be with you. In Lolspeak, the phrase I think so has been lexicalised in this way into an adverb  meaning ‘I think’ or ‘in my opinion’ (aifinkso mebbe it fell behynde the shelfs, ‘I think maybe it fell behind the shelves’).

Examples such as dey lublublubs u foarebber, ‘they love you forever’ or too oar free daze ov sleepsleepsleep, ‘two or three days of sleep’ illustrate intensifying repetition; again, this is a well known process of language evolution, in this case a feature that is characteristic of pidgin languages.

Typical suffixes in Lolspeak are –mus for –ful or –ous (as in byootimus, above, or dangermus). Many suffixes extend the basic form: some examples are –ity, -full and –ify in obviousity, windowfull and insultify (e.g. she dint wanna insultify himz, ‘she didn’t want to insult him’. These forms, together with the deliberate misspelling seen in all the examples here, are typical of internet varieties more generally.  There are new words, too, like nawt sekkund, ‘first’ (kitteh needz tu reed the bukk nawt sekkund so hur can splain it to U layter, ‘kitten needs to read the book first so she can explain it to you later’).

What kind of people belong to the cheezpeeps community? We cannot be sure, since what users say about their identity is not necessarily truthful; nevertheless, it seems likely from their comments and the user profiles that most are women in their forties or older. They are also native or very fluent speakers of English (70% claimed to be from the USA, 10% from Australia, 7% from the UK and about 5% from Canada (with the rest from Europe, Saudi Arabia and Mexico). This is important: cheezpeeps may be enjoying themselves by playing with their language, but at the same time they are trying to impress their audience by demonstrating their high levels of linguistic skill. As Fiorentini points out, being able to play with language in this way, pushing it as far as possible whilst still being comprehensible, needs a high level of understanding of language.

Fiorentini, Ilaria (2013) ‘Zomg? Dis iz a new language’: The case of Lolspeak. Newcastle Working Papers in Linguistics 19 (1): 90-108.

This summary was written by Jenny Cheshire