Monday 24 June 2013

To Split or to Not Split?: The Split Infinitive Past and Present

To be or not to be?
To be or to not be?

There are so many opinions about how to ‘properly’ use language (just think of the debate that the word scone can incite!).  One grammatical structure that has caused a lot of controversy over time is the split infinitive.  The infinitive is the root part of a verb and in English this is written as to + verb (e.g. to walk, to run, to understand, to think, etc.).  More prescriptive grammarians feel that it is not acceptable to split the to and the verb, for example with an adverb, such as to fully understand or with the word not, as in to not allow; in fact many prescriptive grammarians have gone as far as to forbid this practice entirely (as opposed to to entirely forbid this practice’, notice!).

The first recorded instance of a strong aversion to the structure appeared in an anonymous letter to the editor of the New England Magazine in 1834, signed by ‘P.’ P explains that s/he doesn’t know of any rules against the split infinitive but dislikes it because it is only used by ‘uneducated persons’ and in ‘newspapers where the editors have not had the advantage of a good education.’  This idea that infinitives should not be split continues to the present day with The Cambridge Guide to English Usage stating ‘Don’t split an infinitive if the result is an inelegant sentence’ as recently as 2004.

Moises D. Perales-Escudero decided to research how the split infinitive was actually being used in modern American English, as opposed to how grammarians feel it should be used.  To do this he used the Corpus of American English or ‘COCA’ which is an online bank of 385 million words.  He investigated the use of the structure in both spoken and written language situations (=registers) and the texts he studied covered the years from 1990 to 2008. 

When he studied the to not vs. not to structure it very quickly became clear that to not is almost never the preferred form, with not to being used much more frequently in all registers.  Although to not is gradually increasing in use, speakers still prefer the not to combination, which adheres to traditional prescriptive views.

When he began to focus on the to + adverb + verb structure, nevertheless, some more interesting patterns emerged, particularly with regards to how split infinitives are used in certain registers.  We might assume that split infinitives would most often be used in spoken language, due to its more informal and less rigid nature and Escudero-Perales did indeed find that some examples, such as to just + verb’, to really + verb and to actually + verb are much more common in spoken registers.  On the other hand, he also found that other examples, such as to effectively + verb and to better + verb, showed strong associations with written academic registers.  In fact, the split infinitive that was used with most frequency in the COCA was to better understand and it was mainly used in the written academic register, which we would probably perceive as the most rigid with regards to language rules.  However, it seems that it is this particular register which has given rise to its own split infinitive forms, especially those made with to better and to effectively. Perales-Escudero suggests that this could be because academics naturally try to achieve better understandings of phenomena or situations and this is reflected in their language – In other words, it makes clearer sense to keep these words together!  

More research in this area is needed but Escudero-Perales’ study has clearly helped us to better understand the modern use of split infinitives…or has it helped us to understand better…?!
Perales-Escudero, Moises D. (2011) To split or to not split: The split infinitive past and present. Journal of English Linguistics 39: 313-334.

This summary was written by Gemma Stoyle

Monday 17 June 2013

How far does language use reflect identity?

In an increasingly globalised and mobile world, migration is now just a way of life for many people.  This has naturally led to more and more diverse settings for language contact and multilingualism.  Chloe Diskin was interested in exploring the idea that a migrant’s identity plays a significant role in language acquisition.  To do this, she decided to study how features of spoken English are acquired and used in spontaneous conversation by non-native speakers (NNS).  Previous research into this area has shown that NNS whose spoken language use most closely resembles that of native speakers (NS) tend to feel more integrated into society. 

Diskin focused on two different uses of the word like in her study.  Firstly, she considered the use of like as a discourse marker.  Discourse markers are used by NS in conversation for a number of reasons but mainly to ease communication between the participants and add ‘hidden meaning’ to their conversation.  For example, like could be used as the speaker searches for the right word, expresses an approximate quantity, introduces an example or explanation, tries to lessen the force or impact of something they’ve said or repairs a ‘false start’ in the conversation.

The second use of like is known as ‘quotative like’ and is a feature which is currently provoking much interest and discussion in the linguistic world as it seems to be undergoing some rapid changes in use.  ‘Quotative’ words introduce speech and are traditionally words like say or tell.  Over the past 10-15 years like has started to dominate as a quotative, a trend that seems to be led by young females.

Diskin wanted to investigate how closely NNS’s use of these two functions of like correlated with that of NS in Ireland.  She took her sample of NS data from a bank of speech recordings made from 1990 to 1994. The recordings were all conversations between young Irish women.  Her NNS data was taken from informal interviews with two migrant women who are now settled in Ireland.  Agnieszka is Polish and 30 years old, has lived in Dublin for six and a half years and is a Marketing Assistant.  She has an advanced level of English and feels very settled in Dublin, considering herself almost more Irish than Polish now.  Mei Hua is Chinese and is also 30 years old.  She works as a nurse and has lived in Dublin for four years.  She is also very settled in Ireland, living with her Irish boyfriend and enjoying the same level of proficiency in English as Agnieszka. However, Mei Hua left her aging parents back in China and at times feels quite guilty about this.  These details proved important for Diskin later on in her study.

The most ‘fashionable’ form of Irish English at the moment is Dublin English.  In this Dublin English, speakers tend to use the discourse marker ‘like’ mostly in the middle of a phrase, such as in but sometimes I’m like kind of....   Diskin found that this was true of her NS data and also of Agnieszka but not of Mei Hua, who mainly used like at the beginning of phrases, in a way that no NS did.  So it seems that Agnieszka is attempting to adopt the newest and most fashionable trends in Irish English in a way that Mei Hua is not.

Quotative like is emerging as an increasingly prominent feature in recent studies of this fashionable Dublin English.  Agnieszka showed a preference for like over other quotative expressions, again showing how she seems to be following recent trends in Irish English.  On the other hand, Mei Hua only used like as a quotative once in all the data and showed a strong preference for the more traditional say.  Interestingly, in Diskin’s NS data there was not one instance of like being used as a quotative word!  This is most likely to do with the time lapse between the NS and NNS data collection, which goes to show just how quickly a language change can spread!

Diskin speculates that Agnieszka’s and Mei Hua’s language use could reflect how settled and integrated they feel in Irish society.  It seems that Agnieszka, who has a very ‘transnational’ approach to life and admits to feeling more Irish than Polish now, is very clearly adopting the newest trends in spoken language as she attempts to integrate as much as possible into Irish life.  However, although Mei Hua claims to feel settled, she is using much more traditional forms of Irish English than Agnieska.  So maybe ‘feeling integrated’ is not a significant factor in how far NNS acquire spoken speech forms?  Or maybe, whilst Mei Hua appears to be settled in Ireland on the surface, in fact she will never feel rooted there in the same way as Agnieszka, due to her constant unease and guilt about her parents growing old without her in her native China?

This is a fascinating area of research as it proves how deeply entwined our sense of self is with our language acquisition and use.  It definitely lends itself to further research in communities where migrants live and work from day to day.   
Diskin, Chloë (2013) Integration and identity: acquisition of Irish-English by Polish and Chinese migrants in Dublin, Ireland. Newcastle Working Papers in Linguistics 19.1: 67-89.

This summary was written by Gemma Stoyle

Monday 10 June 2013

Revisiting New York department stores

- excuse me, where are ladies's shoes?
- fourth floor!

Patrick-André Mather has recently replicated one of the all-time best known studies in sociolinguistics: William Labov’s classic New York City department store survey.
Records show that as late as the 1890s New Yorkers did not pronounce [r] in words like car or guard. By 1962 though, when Labov carried out his famous survey, pronunciation in New York had begun to change. People now sometimes pronounced [r] in these positions, as in Midwestern American English.
Labov chose three stores that differed in their price range and clientele: Saks Fifth Avenue (the highest ranking store), Klein’s (a bargain basement store) and Macy’s (a middle-ranking store). In each store Labov asked sales assistants where he could find something that he knew was located on the fourth floor. In this way he elicited two words where [r] might be pronounced: once before a consonant (fourth) and once at the end of a word (floor). He then pretended not to have heard the reply, so that the assistants then repeated the same words in a more careful and emphatic speech style. He found that the pronunciation of [r] in the three stores (and by implication in New York City more generally) varied systematically with the social status of the store: assistants in Saks used [r] most often, those in Macy’s used [r] less often and those in Klein’s hardly used [r] at all. The pronunciation of [r] also varied with style, with assistants in all three stories pronouncing [r] more often when asked to repeat the words.
What, then, did Mather find when he repeated the study some five decades later? First, the overall frequency of postvocalic [r] had increased dramatically since the 1960s. This was the case for sales assistants in every store but especially for the high-end store Saks: here floor in the more careful style was now pronounced with [r] 80 per cent of the time, compared to 60 per cent in 1962. The sound change, then, is almost complete at the upper end of the social hierarchy.
Second, the age distribution had changed in Macy’s, the middle-ranking, lower-middle class store. In 1962 younger assistants in Saks used [r] more frequently than older speakers, an age difference that you would expect for a language feature that is undergoing change. In Macy’s, though, it had been older speakers rather than younger speakers who used the new form more frequently, presumably because members of the lower middle class only become aware of a new prestige form as they grow older and their social contacts and social awareness expand. By 2009 the change was now fully underway, and younger speakers in all stores, now including Macy’s, were using [r] more often than older speakers.
There was also an intriguing difference in the use of [r] by African American employees. In the new study both African American and white sales assistants had the same pattern of social and stylistic differentiation, but African Americans used [r] less frequently overall: for them the rate was 50 per cent at Macy’s compared to 60 per cent for all speakers, and under 70 per cent at Saks compared to 80 per cent at Saks for all speakers). On the other hand, African Americans were more sensitive to the phonetic environment of the feature: although all speakers used [r] more frequently in word final position (in floor) than before a consonant (in fourth), the difference was greater for the African American speakers than for the other speakers – African American employees used [r] more than twice as often in floor than in fourth. Mather comments that this pattern of use allows African Americans to maintain a distinct identity whilst still taking part in the general shift towards [r] pronunciation in New York City.
Mather points to a few drawbacks in his ‘trend’ study (in other words, a study of language change where different speakers are sampled at different times within the same community). One drawback was that the lowest level store, Klein’s, had closed down in the 1970s, so he had to find a substitute – and the substitute store needed to have a fourth floor! He chose Filene’s Basement, located close to the original Klein’s store, and also Loehmann’s, needing two lower end stores this time since neither employed as many assistants as Macy’s and Saks. Another issue was that although in 1962 most of the sales assistants were white, by 2009 most of the employees at Macy’s and the two working class stores were African American or Hispanic. What matters, though, Mather argues, is that the sales assistants are representative of the local community in New York City and therefore of the English used in that community, even if the ethnolinguistic makeup of the community has changed over time.
Patrick-André Mather (2012) The social stratification of /r/ in New York City: Labov’s department store study revisited. Journal of English Linguistics 40 (4): 338-365.
doi: 10.1177/0075424211431265  
This summary was written by Jenny Cheshire

Monday 3 June 2013

Quoting then and now

I was like “they’re coming at eleven o’clock “
I said “they’re coming at eleven o’clock” 

Do you use BE LIKE to report what someone said? Thirty years ago few people had heard be like used this way. For young English speakers today, though, BE LIKE has taken over from SAY as the most frequent quotative form. This means that researchers interested in how language changes spread through a language can compare its use by different generations of speakers.

Mercedes Durham and her colleagues note that the most detailed research of this kind comes from Canada. Researchers there have found a strong sex difference emerging as the frequency of BE LIKE increases, with younger generations of female speakers using the new form more often than male speakers. They also found that the kind of quote that BE LIKE introduces changes over the generations:  the first uses of quotative BE LIKE were with a sound indicating the speaker’s state of mind, as in I was like “ugh”, but it was soon also used to introduce reported thought (what someone was thinking), as in I was like “never again”. Only with later generations of speakers is BE LIKE used more often to introduce what someone said (direct speech) than what someone was thinking. 

Durham and her research team analysed the quotative forms used by different generations of undergraduates at the University of York in the UK, to see whether BE LIKE has followed the same pathways of change in York as in Canada. As in Canada, in York the frequency of BE LIKE had soared in just one generation of speakers. In Canada BE LIKE represented 13 per cent of the different quotative forms used by students in 1995; by 2003, the proportion had soared to 63 per cent. In York, too, there was a dramatic increase in the frequency of BE LIKE across the generations, from 19 per cent in 1996 to 68 per cent in 2006. In both locations, then, BE LIKE had taken over from SAY and other quotative verbs to become the most popular quotative form.

However, these figures hide different trajectories of language change. Unlike Canada, in York the difference in the use of BE LIKE by female and male speakers had decreased between 1996 and 2006 rather than increased.  And in York, students in both 1996 and 2006 used BE LIKE the same way – slightly more often to introduce reported thought than direct speech. Across the generations they also continued to use BE LIKE more often with first person subjects and more often in the present tense. In both 1996 and 2006 students in York used SAY and other quotative verbs more often in the past tense.

The researchers point out that the sex differences between Canada and the UK, though interesting, are unremarkable.  They fit with previous research showing that as BE LIKE spreads around the English-speaking world it acquires different social meanings that reflect local social contexts. This results in different social and stylistic patterns in the use of the form from one community to another.

The differences between Canada and the UK in the linguistic effects on the use of BE LIKE are important, though, for our understanding of how changes spread through a language. What has happened in York is consistent with the findings of many researchers working on other kinds of syntactic change in the history of English: successive generations may use a new form more frequently, but they continue to use it in the same linguistic contexts. This is known as the constant rate effect: in other words, as different generations of children acquire the form BE LIKE they also learn the linguistic contexts associated with its use.

Durham and her colleagues suggest, then, that what has happened to BE LIKE in Canada is an exception. They predict that as BE LIKE evolves and spreads in other English-speaking communities around the world it will follow similar pathways of change to what was observed in York: people will use it with increasing frequency but the linguistic effects that constrain its use will remain the same.

This sets an intriguing challenge, then, for researchers elsewhere in the English-speaking world – we’re like “we want to know what happens to BE LIKE”!

To listen to sound clips featuring BE LIKE and other quotatives go to our English Language Teaching Resources website  
 Durham, Mercedes, Haddican, Bill, Zweig, Eytan, Johnson, Daniel Ezra, Baker, Zipporah, Cockeram, David, Danks, Esther and Tyler, Louise (2012) Constant linguistic effects in the diffusion of Be Like. Journal of English Linguistics 40 (4): 316-337.

This summary was written by Jenny Cheshire