This investigation aims to look at the phonological variants of the (str-) consonant cluster in the word initial position in Colchester English. In some dialects 1 , the standard (str-) pronunciation, [stɹ], has undergone a phonological change of palatalisation, whereby the alveolar fricative /s/ anticipates the retroflex nature of /ɹ/, and is thus pronounced with a palato alveolar fricative: [ʃtɹ]. In an article on this phonological phenomenon, Michael Shapiro concludes that the palatalisation of (str-) is an example of assimilation from a distance. This is because the /s/ acquires qualities of the final consonant in the cluster, not the /t/, which remains the same: by stipulating that since /t/ remains /t/ and can, therefore, not figure in the assimilation of /s/ to /r/…it is an example of assimilation at a distance” (Shapiro, 1995: 103). Palatalisation of (str-) has been noticed in a variety of places around the world, including London, where it is currently believed to be affecting the working and middle social classes (Altendorf, 2003).

This paper will investigate whether (str-) palatalisation is a feature that has spread from London into Colchester English. Being situated between London and East Anglia, Colchester is an interesting area to study. This investigation was based on the hypothesis that, as a consequence of counterurbanisation and the increasingly mobile workforce, due to economic liberalisation and improving infrastructure, there will be a spread of dialect features from London, Britain’s economic powerhouse, to surrounding places in the South East of England. A great deal of evidence exists that demonstrates diffusion of phonological innovations from London to its surrounding areas, indeed it has been claimed that London is the source of many phonetic innovations (Foulkes and Docherty, 1999). In fact, diffusion from London to areas like Colchester and other East Anglia population hubs (such as Clacton and Chelmsford) is so likely that Trudgill comments: “the older speakers sound like East Anglians... younger speakers, as is often noted by lay observers, sound like Londoners” (1986: 44). The investigation will focus on comparing the speech of young and old Colchester inhabitants, and will conclude that the palatalised variant is used more frequently by speakers from the younger generation. The investigation will also look at gender differences, to see if the stereotype that women use more standard variants than men is true for the (str-) variable, and to see which gender is leading the language change.


The application of a rapid anonymous survey in Colchester was a relatively simple and efficient process. In order to elicit the desired consonant cluster (str-) in the word initial position, we had to design a set of questions that could have only a limited number of responses; namely a response which would contain at least one variant of the linguistic feature in question – either standard [stɹ] or non-standard palatalized [ʃtɹ]. Upon researching the variable, it was discovered that Labov had partaken in a rapid anonymous survey of the (str-) variable in Philadelphia, USA (Labov, 1984). In it he asked the simple question “Can you tell me how to get to X Avenue?” to which the respondent would reply “Do you mean X Street?” due to the fact the road in question actually had ‘Street’ in the title. It was decided that this efficient method would be implemented, with alterations, in the study of (str-) palatalisation in Colchester English.

Our study was centred on three roads in the town centre of Colchester: the High Street, Head Street and Culver Street. The interviewer, posing as a lost visitor to Colchester, approached the subject on any one of the three streets and firstly asked: “Excuse me, do you know the name of this road?” to which the normal reply would be one of the three aforementioned road names. The interviewer would then ask a secondary question that involved asking for directions to landmarks in Colchester, the Odeon cinema, located on Head Street; and the Colchester department store Williams and Griffin, situated on the High Street. For example, if standing on Head Street or Culver Street, the interviewer would ask “Do you know the way to William and Griffins?” (on the High Street), to which it was hoped the respondent would reply with the directions to the store and the road name where the store was situated: “Turn right onto the High Street”, for example. This way, it was hoped that the questions would elicit two responses that contained an example of at least one variant of the (str-) variable. As the experiment was attempting to test the hypothesis that the (str-) variable is undergoing language change in Colchester English, it was decided that the study should focus on two distinct age groups – distinctly young (aged between sixteen and twenty one) and distinctly old (aged sixty five plus). This decision was based in accordance with the apparent time hypothesis, which states that people from different generations can represent how language was spoken at a particular time. Thus, by comparing two generations that are very far apart, one should be able to observe whether language change is in progress: comparing the generations “allows the researcher to draw diachronic inferences about developments over the last fifty or so years” (Milroy and Gordon, 2003: 35). The experiment was started at around one o’clock; a good time given that the streets of Colchester were filled mostly with students on lunch break from the local sixth form college and elderly retired people spending the day shopping.

Thus, the information recorded in the researcher’s notebook contained (str-) variable used in first answer, (str-) variable used in second answer, estimated age (‘young’ or ‘old’) and the gender of the speaker. Obviously, these notes were not made in front of the speaker as this would make it somewhat obvious that the interviewer had an ulterior motive. Rather, the interviewer would continue along the road for a few metres before making the notes. In order to see if this method would be successful, a preliminary testing stage was used on ten subjects.

Early Methodological Problems

It quickly became apparent in the testing stage that the first question asked – “Which road is this?” would be a successful way to elicit the response of either “ X [stɹi:t]” or “ X [ʃtɹi:t]”. Only in the minority of cases did the response not involve the word “Street”, namely when the person asked was not actually aware of the name of the road, perhaps due to the fact they were not actually from Colchester. Only the results that did contain a variant of the (str-) variable in the word “street” were recorded. However, there was a problem with the data collection process, namely it was not always possible to elicit a second response containing the word “street”. Indeed, sometimes a second “street” was uttered without the prompting of another question – many people uttered “I think its Head Street, yes, Head Street”, as if to confirm their first answer. However, when the interviewer had to ask the second planned question, “Can you direct me to the Odeon/William and Griffin?” the subject would often respond with solely directions, and not mention the road name. In these cases, the interviewer tried once more to elicit the word “street” by asking “Do you know what road it is on?” – a question which received mixed results. Sometimes a second “street” was uttered; however, many times, the subject would just look confused as to why this question was asked and instead just repeat the directions they had already given. Perhaps the third question was one step too far; bearing in mind the aim of a rapid anonymous survey is to ask an unobtrusive question, it is likely the subjects were uncomfortable being probed for answers by a stranger, and perhaps annoyed at the interruption to their day.

However, in general the method was successful in gaining at least one response that included the word “street”. Some secondary responses were elicited, but not as many as hoped. These secondary responses were still recorded, and will be discussed below. In total, one hundred and twenty subjects were interviewed in one hour and a half. Of these, one hundred and one used the word “street” in their first response. Only these subjects were recorded. Of the one hundred and one recorded subjects, thirty eight uttered “street” a second time. The subjects were also well balanced between young and old (fifty seven young, forty four old) and between male and female (fifty one male and fifty female). To collate the results, a percentage use of non-standard variant [ʃtɹ] was calculated for both the first responses and the second responses. This was achieved by dividing the number of non-standard variants used by the number of (str-) variable occurrences, and multiplying by one hundred. The methodology allowed the interviewer to gain many examples of two different variants in the dialect: “the exchange was a systematic elicitation of the exact forms required, in the desired context… with the desired contrast of style” (Labov, 1972: 310), whilst the subject was oblivious to their participation in a linguistic experiment. Rather, they assumed they were responding to a genuine question from a lost and confused stranger, thus their language use should not have been affected: the “tenuous relationship was the minimum intrusion upon the behaviour of the subject” (ibid).

Results and Analysis

Table 1: Comparing the first responses of young subjects to the first responses of old subjects

No. of [stɹ]

variant used

No. of [ʃtɹ] variant used

Total number

of (str-)


% use of non-standard [ʃtɹ]
Young 18 39 57



Old 33 11 44



Total 51 50 101



These results show a clear difference between the speech of those classed as distinctly young and those classed as distinctly old. Sixty eight percent of the young age group used the non-standard palatalised variant [ʃtɹ] from London, the majority of that age group, whereas seventy five percent of the older age group used the non-palatalised variant [stɹ] that is standard to Colchester English. The use of variable (str-) could therefore be interpreted as a change in progress, and thus be seen as evidence of a London English feature diffusing into Colchester English. It could be the case that if the apparent diffusion of the palatalised variant continues, the non-palatalised variant could die out in Colchester English, although currently is still in use. These assumptions are again based on the aforementioned apparent time hypothesis (Milroy and Gordon, 2003).

Table 2:Comparing the first responses of male subjects to the first responses of female subjects

No. of [stɹ] variant used No. of [ʃtɹ] variant used

Total number

of (str-)


% use of non-

standard [ʃtɹ]

Men 21 30 51



Women 30 20 50



Total 51 50 101



Table 2 shows that the variant used is also influenced by the gender of the speaker. As aforementioned, gender is a highly debated topic in variation studies. It would be futile to suggest that the results from table two show that women use a more standard form than men. However, in the restricted context of the word initial consonant cluster (str-) in Colchester English, it would appear that more men use the non-standard palatalised variant [ʃtɹ] than women. This pattern is perhaps unexpected given the belief that, broadly speaking, women are said to lead linguistic change 2 . Therefore, perhaps one should expect to find women using the [ʃtɹ] variant more than men if this feature was really undergoing linguistic change in Colchester English. However, the results from table 1 and table 3, which will be discussed below, show a clear difference in the variant of (str-) used between the young and old. This would suggest language change is in progress in Colchester, with regard to the (str-) variable in word initial positions. Perhaps it is the case that, in this situation, the linguistic change is being led by males. Indeed, despite being in the minority, some linguistic changes that have been recorded have been led by men; see for example Labov (1963).

Table 3:– Comparing the first responses of young male subjects to the first responses of old male subjects, and young female subjects to old female subjects

No. of [stɹ] variant used No. of [ʃtɹ] variant used

Total number

of (str-)


% use of



Young Men 7 22 29



Young Women 11 17 28



Old Men 14 8 22



Old Women 19 3 22



Total 51 50 101



Table 4 below presents a graphical version of the results above, and a commentary follows.

Table 4:Graph charting the percentage use of the non-standard variant [ʃtɹ] of the (str-) variable in Colchester English, comparing young male and female first responses to old male and female first responses

After dividing the young and old into male and female, it is evident that the initial patterns observed in tables 1 and 2 remain. A considerably higher percentage of young men use the non-standard palatalised variant [ʃtɹ] compared to old men. Likewise, young women are more likely to use the palatalised variant than old women. In addition, the gender differences are apparent in both age groups, with a lower percentage of young women using the palatalised variant than young men and a lower percentage of old women using the palatalised variant than old men. However, age appears to be the most important factor. Even though young women use the standard non-palatalised variant more than young men, they still use the non-standard palatalised variant [ʃtɹ] more than old men. This would match the analysis of table 1 as possibly demonstrating an ongoing phonological change, despite the gender differences.

Table 5:Second Responses

No. of [stɹ]



No. of




Total number

of (str-)


% use

of non-standard


Young Men 5 9 14



Old Men 5 5 10



Young Women 4 3 7



Old Women 5 2 7



Total 19 19 38



The collection of two responses was not greatly successful, and was hampered by the factors explained above in the methodology section. However, from the small set of second responses involving the word “street” that were collected, some patterns that emerged in the first responses are maintained – a greater proportion of young men use non-standard (str-) than old men, whilst more young women use the non-standard than old women. However, in these responses, a higher percentage of old men use the palatalised variant [ʃtɹ] than young women. This is a strange result that does not keep to the patterns established in the results from the first responses. Unfortunately not enough results were collected for the secondary responses, and therefore it would not be beneficiary to make any sweeping conclusion given these sets of results.

Thus, it would appear that speakers in the younger generation are more likely to use the palatalised variant than those in the older generation. This would suggest, using the aforementioned apparent time hypothesis, that the word initial (str-) variable in Colchester English is undergoing change, with the palatalised variant [ʃtɹ] becoming more common, possibly as a result of diffusion from London English. A problem with the apparent time hypothesis is that a person does not speak in consistently the same way throughout their lives. Rather, particular forms or styles are more commonly used at particular stages in life, a phenomenon known as age-grading: “using speech appropriate to your age group” (Wardhaugh, 2006: 196). This is potentially damaging to this experiment, as it could be the case that the differences observed are simply a consequence of age-grading, and not language change. However, the feature in question is not one of high social awareness or one that is consciously manipulated, and thus it is unlikely to be influenced by the age-grading phenomenon (Milroy and Gordon, 2003).

Evaluation of Experiment Methodology

Firstly, one must point to some criticisms of the methodology. As previously examined, not enough secondary responses were elicited to enable them to be evaluated seriously. It would have been interesting to see if the strange patterns that emerged in the second responses (see table 5) were maintained had more data been collected. In addition, there are some key independent variables that are missing from the data which, with hindsight, perhaps should have been included. One error was not recording the name of the road the participant uttered. This could have been important in that the preceding sound could have influenced the choice of (str-) variant used. Both Culver Street and the High Street have vowels at the end of the first word: ‘Culver’ could be pronounced with a central vowel [ɜ:] or a schwa [ə]; ‘High’ ends in the diphthong [aɪ]; whilst Head Street has a voiced alveolar stop [d]. It would have been interesting to see if the preceding sound influenced the pronunciation of the following (str-) cluster. In addition, it would perhaps have been interesting to look at some other social factors, such as social class or ethnicity. However, in this type of experiment, it would be practically impossible to guess a participant’s social class from such a short exchange. Perhaps the experiment could have been carried out in two places with different inhabitants or two shops with different target shoppers (as Labov did in his famous survey of New York department stores (1972)).

It is also necessary to point to the limitations of this type of experiment. A major problem is the data being unrecorded, thus the unconscious bias of the interviewer in selecting which variant the subject used may have played a part, distorting the answers: “it is always possible that an unconscious bias in transcription would led to some doubtful cases being recorded” (Labov, 1972: 323). In addition, the experiment could be made more fair and systematic by picking out only a certain number of random subjects, rather than choosing which people to ask. For example, every fifth person could have been asked to make the experiment more random. The rapid anonymous survey methodology gives the researcher a sample of the population that is present on the street (Labov, 1984); it is unknown how close this representation is to the total population of Colchester. Furthermore, one cannot be certain that all subjects were Colchester residents, as to ask where they reside would have violated the anonymity of the surveyor. To verify the results, perhaps a more detailed study could be embarked upon, for example a series of sociolinguistic interviews could be made, whereby many casual conversations are recorded and transcribed.

However, despite the problems with the rapid anonymous survey method, there are some very large advantages. The experiment was quick, achieving one hundred and one examples of a phonological variable that would usually be relatively infrequent in a person’s speech in just one hour and a half. In addition, the experiment shows a clear pattern of results with speech differences between young subjects and old subjects, as was hypothesised.


To conclude, this rapid anonymous survey has highlighted key differences in the pronunciation of the word initial (str-) variable in Colchester English. It would appear that language change is in progress, given the non-standard palatalised variant was more likely to be used by the participants from the younger generation. This is possibly as a result of the feature diffusing from London English. In addition, it would appear that male subjects are more likely to use the palatalised variant than women, which could suggest that, unusually, in this case, linguistic change is being led by men.


Altendorf, U. (2003): Estuary English: levelling at the interface of RP and South-Eastern British English. Tübingen: Narr.

Britain, D. (2008): Dialect Variation: an introduction to language variability. LG254 Week 5 Lecture Notes. University of Essex.

Cedergren, H. (1973): The interplay of social and linguistic factors in Panama. Unpublished Cornell University dissertation.

Chambers, J.K. (1995): Sociolinguistic Theory: Linguistic variation and its social significance. Oxford: Blackwell.

Eckert, P. (1986): The roles of high school social structure in phonological change . Chicago Linguistics Society.

Fasold, R. (1969): A sociolinguistic study of the pronunciation of three vowels in Detroit speech . Mimeograph, Center for Applied Linguistics, 1968.

Harrison, S. (1999): ‘English /(s)tr/ clusters’. LINGUIST List 10.217 , online at .

Labov, W. (1963): ‘The social motivation of a sound change’. Word 19 , pp. 273-309.

Labov, W. (1966): The Social Stratification of English in New York City . Washington, DC: Center for Applied Linguistics.

Labov, W. (1972): ‘ The Social Stratification of (R) in New York City Department Stores.’ In H. Allen, and D. Linn (eds.) (1986) Dialect and Language Variation. London: Academic Press.

Labov, W. (1984): ‘ Field methods of the project on linguistic change and variation.’ In J. Baugh, and J. Sherzer (eds.) Language in Use: Readings in Sociolinguistics. Englewood Cliffs, NJ: Prentice Hall, pp.28-66.

Luthin, H. (1987): The story of California (ow): the coming-of-age of English in California. In Denning, K. et al. (eds.) Variation in Language: NWAV-XV at Stanford : Department of Linguistics, Stanford University, pp. 312-24.

Milroy, L. and Gordon, M. (2003): Sociolinguistics Method and Interpretation. Oxford: Blackwell Publishing.

Shapiro, M. (1995): ‘ A case of distant assimilation: /str/ - /ʃtr/. American Speech, 70 (1) , pp. 101-107. Duke University Press.

Trudgill, P. (1974): The Social Differentiation of English in Norwich . Cambridge: CUP.

Trudgill, P. (1986): Dialects in contact . Oxford: Blackwell.

Wardhaugh, R. (2006). An Introduction to Sociolinguistics Fifth Edition. Oxford: Blackwell Publishing.

©Michael Bass. This article is licensed under a Creative Commons Attribution 4.0 International Licence (CC BY).

  1. (str-) palatalisation is currently observed in London English, and is thought to be spreading to other areas in the South East of England (Harrison, 1999), for example, Canterbury in Kent (Altendorf, 2003: 101).

  2. Women are shown to be the innovators of linguistic change in most recent studies, with new innovators appearing in the speech of women before men (Labov, 1966; Trudgill, 1974; Cerdergren, 1973; Fasold, 1969; Eckert, 1986; Luthin, 1987). This is perhaps due to women’s superiority to men in most aspects of verbal behaviour (Chambers, 1995: 291).