The material in this page, which is presented in an extremely condensed form is drawn entirely from Jones (2002). Herein, consideration is given solely to the evolution of definite article reduction (DAR). The development of the definite article in Standard British English is discussed in the Overview section. All references to "localities" within the text relate to survey points within northern England and not solely to those within the boundaries of Yorkshire.

In relation to the opening quote, Jones both examines the notion of assimilation and tests the theory using material from Ellis (1889) and the Survey of English Dialects (SED). It should be noted that all citations and examples given below are attributable to Jones.

progressive assimilation: refers to the modification of a speech sound so that it becomes more like the sound preceding it. Thus, in this case, the dental fricative [] assimilates to the alveolar plosives of [t] and [d]

fricative: sound produced by bringing two articulators (e.g. tongue and upper teeth) together but not close enough to stop the airflow (thus creating friction), e.g. th as in thin which is denoted phonetically as [].

elision: the omission of a sound or sounds in speech. For example library [] is sometimes pronounced as []

Assimilation theory

According to Jones, a general consensus exists that all the various forms of DAR have their origins in progressive assimilation of the fricative of Middle English (ME) þe. The resulting form te is well attested in a number of ME texts. The accepted thinking is that this assimilated form marks the primary phase in the development of the reduced article.

However, it should be noted that the processes of assimilation occurred in a large number of ME dialects and not just in those currently featuring DAR. Moreover, words other than the definite article were also affected e.g.

þe	they	þ	thou
þeggre	their	þær	there
þiss	this	þærfore	therefore
þohh	though	þanne	then

The vowel-less forms of the reduced article identified in the SED (e.g. [], [], [] etc) are believed to be the result of elision. However, Viereck (1995: 299) opines that such elision may only have happened following the modified consonantal realization mentioned above. Jones supports this notion by pointing out that there is no textual evidence of non-fricative vowel-less forms (i.e. generally written as t' in dialect literature) prior to George Meriton's A Yorkshire Dialogue in 1673. He also notes that the elision of the vowel in fricative forms prior to this date are seen to occur only when closely located to vowels.

The Conditioning environment

Examination of different ME texts (e.g. The Bestiary, The Ormulum) reveals that te appears to a large extent after /t/ and /d/ (e.g. þatt te "that the" and mid te "with the" - see Vierick, 1995). Other occurrences of te have also been found following /s/ (Reaney, 1927) though Jones expresses uncertainty as to whether this is a "geographical or diachronic variation". However, the idea that such phonological environments are related to the evolution of DAR runs up against a number of problems, mainly that:

• In some dialects, whilst other words (e.g. þe ) underwent the assimilatory change, the definite article remained unaffected.

• Given that te was so widespread in ME dialects, some explanation is required as to why this form was maintained only in the DAR dialects.

Jones (2000) reasons that if the assimilation theory described above is indeed the source of DAR, then present-day variations should not take the fricative form when immediately preceded by /t/, /d/ or /s/ (and, he suggests, word-final /z/). To test this theory (henceforth known as the development hypothesis) he examines data from Ellis (1889) and the SED.

The Data

The Ellis data
This data is derived from two source texts which have been transcribed into dialect, namely: the comparative specimen and the dialect test. Jones examines each of these datasets separately on the grounds that, as a general rule, only one text per locality was employed for the purposes of data collection. In addition, as observed by Jones, the "phonological environments including the definite article are not comparable between the two". His findings are as follows:

The comparative specimen

Of the 64 localities included in the database only 24 reveal examples of DAR in both fricative and non-fricative forms. Analysis of this sample reveals that:

• The /t/ environment occurs in the majority of localities most of which show DAR in the fricative form.

• The /s/ environment is found at a number of the survey points and mainly precedes a fricative.

• Of the two occurrences of the /d/ environment one is followed by a fricative.
  
  

Thus the majority of these environments display a fricative form of the reduced article (e.g. at th' corner, that th' youngest) and accordingly the development hypothesis is not supported.

The dialect test

Although the data includes in excess of 300 occurrences of DAR, the source text contains only a single phrase (i.e. won't the old) incorporating the phonological environment responsible for the activation of the assimilatory processes recognized in Middle English texts. Jones identifies only six locations displaying allomorphs possibly relating to the development of DAR. He includes a further six localities on the basis of (t') assimilating to (s) or (s'). Within these samples the data from only a single locality provides support for the development hypothesis.

The SED data
In relation to the occurrence of DAR, Jones examines all the data in the Basic Material including that from the incidental material. He also analyses the actual recordings. Forms of "the one" and "the other" are universally disregarded as being of possibly different historical origins (see, for example, the suggested origins of tother). The findings are as follows:

Basic material

Jones identifies 2,621 examples of DAR over 101 localities. However, after the exclusion of cases displaying linguistic environmental ambiguity and localities deemed to show insufficient or irrelevant data, only 39 locations exhibiting variability between fricative and non-fricative forms remain for analysis. In contrast to the findings from the Ellis material, the data from 22 of these survey points support the development hypothesis.

Recordings

Within this dataset 2,228 instances of DAR are identified although some survey points are ignored because of the small numbers of reduced articles in evidence. However, only 17 localities demonstrated variability between fricatives and non-fricatives and only 9 of these support the development hypothesis.

The Overall Position
Taking into account the findings from both sets of data, support for the development hypothesis appears to be relatively sparse. The overall position is summarised in the following map.

Conclusion

Jones holds the view that a further more detailed wide-ranging survey is required before an alternative model of DAR development can be formulated. However, he does offer some speculation on the origins of the zero form in the Holderness region of the East Riding of Yorkshire. He also puts forward his own theory regarding the development of DAR.

In relation to the zero forms of the Holderness area, Jones calls into question the proposal that this development links back to the time of Norse settlement. Discounting the claim that Old Danish gave rise to the zero form whilst plosive forms have their roots in Old Norwegian. Jones observes that, whilst [] > [] might be considered to be a natural phonological progression, the East Riding displays "no linguistic separation from the adjoining dialect area in other major respects (see Wakelin. 1977: 102-3)" He therefore takes the view that it would be an extraordinary occurrence if the definite article alone was influenced by the effects of linguistic contact.

Concerning DAR as a whole, Jones states that there is no evidence to support the notion that the origin lies in the final <t> of the Old English demonstrative þat. He observes that, judging from the orthographic convention employed in A Yorkshire Dialogue (Cawley, 1959), the present-day forms of allomorphic variation appear to have evolved by the late 17th century. Furthermore, all such variations consist fundamentally of three sounds, [t], [] and [] in different combinations (see Variations). Because of known crosslinguistic mutations (i.e. // > /t/ and /t/ > //) Jones suggests that it is these three basic sounds which are the source of DAR.

Sources

Cawley, A.C. (ed) (1959) George Meriton's 'A Yorkshire Dialogue'. 1683. Yorkshire Dialect Society, reprint 2.

Ellis, A. J. (1889) On Early English Pronunciation, Part V. The existing phonology of English dialects compared with that of West Saxon speech. London: Trübner and Co.

Jones, M. J. (2002) The origin of Definite Article Reduction in northern English dialects: evidence from dialect allomorphy. English Language and Linguistics 6.2: 325-345.

Reaney, P. H. (1927) A grammar of the dialect of Penrith (Cumberland) Publications of the University of Manchester, English series 15.

Viereck, W. (1995) Realizations of the definite article in dialectal English and how and when they originated. In Fisiak, J. (ed) Medieval dialectology, The Hague: Mouton de Gruyter. 295-307.

Wakelin, M. F. (1977) English Dialects: an introduction, London: Athlone Press.