Relativiser that in Scottish English news writing

Sanna Hillberg
University of Eastern Finland


Relativisation is a well-researched area of English syntax. However, the main interest has been on spoken language and investigations on relativisation strategies in written language are scarce. My forthcoming PhD thesis seeks to fill in this gap. In this paper I discuss some preliminary findings of my work with respect to the relativiser that use in written news texts in Highland and Island Scottish English (HIScE) and Lowland Scottish English (LScE). These findings are compared with those in Standard English and Irish English.

As my database I use the Corpus of Scottish English On-line Press News (CSEOPN), which is a corpus I have compiled for the purposes of my PhD study, and the written news reportage sections of two components of the widely-used International Corpus of English (ICE), viz. ICE-Great Britain Release 2 and ICE-Ireland.

The findings show that differences arise in the relativiser that use between regional varieties of educated English of the British Isles, suggesting that variation in relativisation is not “an exclusive right” of spoken dialects, but extends to standard written varieties as well.

1. Introduction

This paper discusses features of the relativiser that use in written Educated Scottish Standard English (ESSE), and the data is drawn from Scottish newspapers which are published online. [1] Scottish English (ScE) is the northernmost variety of English in the British Isles. Features of Scottish English grammar have been discussed, for example, by Miller (1993, 2008), Bergs (2001) and Macafee (2011), while Macafee (1983) and Macaulay (1991) have described Lowland Scots dialects of Glasgow and Ayr in detail, respectively. Relativisation in Lowland Scots dialects has been documented recently by Herrmann (2005) and Tagliamonte et al. (2005). Romaine (1980, 1982, 1985), who is often cited in articles discussing relativisation, has investigated relativisation in Middle Scots writings and in contemporary spoken Scots, and she also briefly discusses relativiser use in written ScE. All the abovementioned works on present-day Lowland Scots concentrate on its spoken form. However, relativisation strategies in Highland and Island Scottish English (HIScE) and in written modern ScE have successfully evaded scholarly investigation so far. With this paper I hope to cast light on the relativisation strategies of these varieties.

Terminology with regard to languages spoken in Scotland may be somewhat confusing for those not familiar with the area. Scots is a dialect / language spoken and written mainly in the Lowlands of Scotland. It is of considerable debate whether Scots can be regarded as a language of its own right or whether it is a dialect of English (see e.g. Macafee 1996; Douglas 2009). Without taking sides in this matter, it can be said that Scots displays distinct vocabulary and syntactic features. Scottish English is English spoken in Scotland with a Scottish accent. These accents come in many different forms and are described, for example, by Aitken (1984a). Scottish Gaelic, on the other hand, is a Celtic language and not related to English. It was once the dominant language in the Western Isles, where it still has some native speakers, and Western and Northern Scotland.

Most studies concerning relativiser use discuss the use of restrictive relatives only (see e.g. Tottie 1997; Beal and Corrigan 2002; Tagliamonte et al. 2005; Huber 2012). This has been explained by the overwhelming amount of wh-relatives in comparison to the zero relativiser and that in non-restrictive relative clauses (see e.g. Ball 1996). While this is true, Herrmann (2003: 107), however, shows that non-restrictive that is a feature of dialectal speech and accounts for seven per cent of all non-restrictive relative clauses in her study of six spoken varieties of British English. As shown by Biber et al. (1999: 610) a small amount of non-restrictive that is found in BrE and AmE written news texts. Therefore, non-restrictive relative clauses should not be disregarded in studies that concentrate on relativisation.

In this study evidence from restrictive and non-restrictive that has been merged. Because I will not compare the use of the relativiser that to other relatives, I do not find it mandatory to distinguish between restrictive and non-restrictive. My aim is to provide a general overview of relativiser that functions in ESSE, and it will be left for my future studies to analyse between restrictive and non-restrictive that in more detail.

Another aim of this paper is to find out whether variation exists within standard written Englishes of the British Isles. As Biber et al. (1999: 18–19) point out, differences are manifested between the standards of BrE and AmE on many linguistic levels, for example, in the use of relative clauses. Nevertheless, variation in written StE has not received much scholarly attention so far. The study of standard written language may have been delayed because of earlier notions on the standard being the same around the British Isles. For example, Petyt (1980: 14–15) claims that speakers of different English dialects “find the same written form in their local newspapers”. It is unclear whether ‘form’ here means simply orthographical similarity or includes grammar as well. The hypothesis for the current study was that variation between the ESSE varieties with respect to the relativiser that use would be found, because the linguistic and socio-historical backgrounds of Highlands and Lowlands of Scotland are very different from each other (see e.g. Görlach 2002; Shuken 1984).

It is important to highlight here that this paper discusses English that is used and published in Scotland. As I was collecting my database it was impossible to make sure that all journalists whose work is included in my data of ESSE fulfil the strict standards set for sociolinguistic studies, i.e. whether or not these writers are born and bred in Scotland and native ScE speakers. The data, however, represents the actual language use in the region, and these news have been written for and read by the local Scottish audience.

The structure of this paper is the following: I will first briefly discuss the history of the relativiser that in Lowland Scots, after which I will introduce the data and methods of my research. In section four the preliminary results of the study will be elaborated, and the last section is reserved for conclusions and discussion.

2. A brief history of the relativiser that in Scots

The indeclinable Older Scots relativiser had two forms: that and at. [2] It is debated whether at is merely a corruption of that or derives from Scandinavian (Caldwell 1974: 31; King 1997: 172). Caldwell, however, considers these two forms separately in her study of relativisation strategies in Early Scots writings (1375–1500), and reports that the instances of the relativiser that clearly outnumber those of the relativiser at. The use of at declines after the sixteenth century in written Scots. Caldwell (1974: 31) claims that the relativiser and conjunction at, nowadays most likely a “worn-down” variant of that, may still be heard in spoken Scots and the Penrith dialect of northern English (see also Millar 2007: 70).

In Older Scots the relativiser that functions freely in restrictive and non-restrictive relative clauses, while in Present-Day Standard English (PdE) it is typically accepted only in restrictive relative clauses (see, e.g. Quirk et al. 1985: 366; Huddleston and Pullum 2002: 1059). Caldwell’s (1974: 27–8, 72–3) data shows interesting distinction in the uses of restrictive and non-restrictive that. In literary texts that is frequent in both types of adnominal relative clauses, while in record and official prose the relativiser that is common in restrictive relative clauses, but infrequent in non-restrictive relative clauses. At the same time, (the) quhilk(is) (precursor of the relativiser which) emerges as a non-restrictive relativiser. These findings suggest that differences in the uses of that and (the) quhilk(is) with respect to restrictiveness as well as limitations to particular register use emerge during the ES period: that starts to lose ground as a non-restrictive relativiser in formal written discourse, and becomes more pronounced as a relativiser of spoken language. Gradually, (the) quhilk(is) becomes the norm of non-restrictive relative clauses in formal writing, but does not really infiltrate the spoken system (see e.g. Romaine 1980; Tagliamonte 2005). Caldwell’s (1974) findings therefore suggest that the relativiser that might have been considered somewhat informal already during the Older Scots period. In Middle Scots texts studied by Romaine (1982) that typically functions in restrictive relative clauses, which parallels its modern StE and ScE use.

Another obvious reason why that became rare in non-restrictive relative clauses is the implementation of the personal relativiser who into Scots in the 16th century. The relativiser who appears in Scots texts approximately a hundred years later than in southern English (Caldwell 1974: 43). Who often modifies proper nouns, which are always followed by a non-restrictive relative clause. Therefore, who was a natural alternative with personal/human antecedents. Also, according to Rydén (1979: 14), that was functionally overloaded and wh-relatives were more “suitable to express exact relation”.

In Older Scots that operates in similar syntactic functions than today, i.e. as a subject, object, adverbial and prepositional and copulative complement (Moessner 1997). Preposition use with the relativiser that/at is uncommon in Caldwell’s data of ES writings. Prepositional that is found typically in restrictive relative clauses, where it follows the verb and the preposition is placed at the end of the relative clause (i.e. preposition is stranded). With respect to the type of the antecedent that functions with human, non-human and inanimate antecedents (King 1997: 172, Romaine 1982). As mentioned above, during the Older Scots period, who started to gain ground as a relativiser of human antecedents in written discourse, albeit the process was slow (see e.g. Meurman-Solin 2000). However, in Modern spoken Scots that is still common with human antecedents (see e.g. Tagliamonte et al. 2005)

Previous research on spoken varieties on Lowland Scots has revealed that its relativisation strategies function rather differently from StE, which is, of course, typical for nonstandard varieties of English (see e.g. Herrmann 2005). For example, in spoken Scots relative clauses that + possessive pronoun may substitute the genitive whose and subject relativiser that can be omitted in existential clauses (Miller 1993: 111–112). Unsurprisingly, these features are completely absent from my material, which is, as mentioned before, collected from written sources.

3. Data and Methods

My database consists of corpora of newspaper articles concerning national and local news. The Corpus of Scottish English On-line Press News (CSEOPN) is compiled by myself, and it contains approximately 172,000 words. The CSEOPN is divided into two sub-corpora: HIScE (86,259 words) and LScE (86,174 words), which have been compiled from newspapers published in sixteen localities in the Highlands and Lowlands of Scotland (for the composition of CSEOPN and a map of HIScE and LScE-speaking areas of Scotland see Hillberg, forthcoming).

As points of comparison I use two components of the International Corpus of English (ICE), viz. ICE-Great Britain Release 2 and ICE-Ireland. However, the articles from Scottish English newspapers in ICE-GB have been replaced by articles from the Guardian and the Times. For clarity, I will refer to the combined data of ICE-GB and the articles from the Guardian and the Times as StE. In my analysis of the data I have used the data retrieval program MonoConc, which helps to find relative clause structures in the corpora. In the case of ICE-GB, I have used its own concordance program, ICECUP, to find the relativiser that instances.

As Tagliamonte et al. (2005) state there are two factors that have “strongly constrained the choice of relative marker since early-modern English”, namely the syntactic function of the relativiser in the relative clause and the type of antecedent. Because of this fact the features that will be discussed in this paper are the following: the type of the adnominal relative clause (restrictive vs. non-restrictive); the grammatical function of that in the relative clause; the gender type of the antecedent (human, inanimate vs. animate); and definiteness of the antecedent (definite vs. indefinite).

It also has to be mentioned that it-cleft clauses have been excluded from the data. Although they function superficially like relative clauses, their underlying function is different. Therefore, the position of it-clefts is somewhat controversial. Some grammarians do not consider them relative clauses on the basis of their structure (see e.g. Quirk et al. 1985; Biber et al. 1999), while others see no problem placing them in the category of relative clauses (Huddleston and Pullum 2002).

4. Relativiser that in ESSE

The wh-relatives are considered more formal than that and zero in written registers (see e.g. Quirk et al 1985; Biber et al. 1999), while that and zero are clearly more frequent in dialectal speech (see e.g. Herrmann 2005; Tagliamonte et al. 2005). As table 1 shows, the wh-relatives which and who dominate the data, while that and the zero relative are less common. Nevertheless, that is frequent in news writing in all varieties under scrutiny except for IrE, which seems to strongly favour wh-relativisers in news writing. The relativiser that is clearly more frequent in LScE (N=142 out of the total of 778 relativisers in the data) in comparison to HIScE (N=115 out of the total of 799 relativisers), and this difference is statistically significant (two-tailed chi-square with Yates’ correction: χ2= 4.025, df = 1, p-value = 0.0448). This result suggests that with respect to the use of the relativiser that HIScE strives towards greater formality in news writing than LScE.

Relativiser which who whom whose that zero total
HIScE 34.4 26.5 0.3 1.9 13.3 16.1 92.6
LScE 29.5 26.9 0.5 1.4 16.5 15.5 90.3
ICE-Ireland 30.4 35.2 0.5 2.3 7.4 9.4 85.1
StE 33.4 25.6 0.5 2.7 14.2 11.4 87.7

Table 1. The occurrence of relativisers per 10,000 words of running text.

Of the four varieties under scrutiny, that is most frequent in LScE, in which it is also the most common relativiser in speech (see e.g. Herrmann 2005; Tagliamonte et al. 2005). It is possible that the frequent use of that in LScE reflects influence from spoken Scots. Distribution of relativisers in the data (see Figure 1 below) show that that is somewhat more frequent in StE than in HIScE, but there is no clear difference (two-tailed chi-square with Yates’ correction: χ2= 0.397, df = 1, p-value = 0.5284). That is rather uncommon in the IrE data, and occurs in less than every tenth relative clause, which is less than a half of the occurrences in comparison to LScE.

Figure 1

Figure 1. Distribution of relativisers in each variety.

4.1 Restrictive vs. nonrestrictive that

Restrictive relative clauses provide information about the antecedent which is necessary for its identification, while the information a non-restrictive relative clause carries is additional, and ‘only loosely related to the surrounding structure’ (Huddleston and Pullum 2002: 1058). Recent studies on relativisation have mainly concentrated on restrictive relative clauses. Ball (1996: 228–9) explains:

Studies on relative clauses are generally careful to distinguish between RRs and non-restrictive relative clauses… [3] The distinction is crucial for quantitative studies of relative markers, because NRRs strongly favour WH-forms in standard English, whereas RRs do so variably. The inclusion of NRRs in a sample therefore tends to raise the percentage of WH-forms and lower the percentages of that and zero.

Reports on nonstandard varieties of English suggest that non-restrictive that is not alien to dialectal speech (see e.g. Herrmann 2005, eWAVE). Beal and Corrigan (2002: 128) mention that “in Tyneside English wh- is not categorical in non-restrictive relative clauses”, which implies that the relativiser that is a feature of this Northern English variety as well.

Figure 2

Figure 2. Restrictive and non-restrictive that in HIScE, LScE, IrE and StE.

Figure 2 shows that non-restrictive that occurs in all varieties, although according to Quirk et al. (1985: 1258) that is “very rare” in non-restrictive relative clauses. Huddleston and Pullum (2002: 1059) state that the acceptability of that in non-restrictive relative clauses is marginal, and wh-relatives are “required or strongly favoured in supplementary relatives” (2002: 1052). [4] In Biber et al.’s (1999) data of AmE and BrE news writing non-restrictive that is infrequent and it does not occur in fiction or academic writing at all (1999: 610–1). They further state that the rare cases of non-restrictive that either occur “in a series of postmodifiers” or it “is used for special stylistic effect (especially in fiction)” (1999: 615). Based on these observations, it was expected that non-restrictive that would not be found in any great quantities in the news articles under investigation.

In the current data the clear majority of that relative clauses are restrictive, but contrary to expectations non-restrictive that is a regular feature of news writing. In the StE data nearly every fourth that relative clause is non-restrictive. The occurrence of non-restrictive that in ESSE is clearly smaller, less than ten per cent in both varieties. The variation between the ESSE varieties and StE yields a statistically significant result (two-tailed chi-square with Yates’ correction: χ2= 12.750, df = 1, p-value = 0.0004). In ICE-Ireland news texts every tenth that relative clause is non-restrictive, which means that this feature is more common in IrE than in ScE news writing.

The difference between restrictive and non-restrictive that between the ESSE varieties is not statistically significant (two-tailed chi-square with Yates’ correction: χ2= 0.039, df = 1, p-value = 0.8428). Proportionally non-restrictive that is slightly more frequent in HIScE (7.0%) in comparison to LScE (6.3%). Earlier studies show that non-restrictive that is found in spoken Lowland Scots (see e.g. Macafee 1983; Herrmann 2003). It is possible that spoken Scots influences non-restrictive that usage in written Scottish news, but it is also equally likely that this usage derives from StE. It also may be possible that Scottish writers follow StE norms in the use of restrictive that more strictly than their colleagues south of the border.

Non-restrictive that (1) typically occurs without a comma in the current data:

(1) The committee agreed to scrap the £86,000-a-year contract with SPPA that has been in place for 10 years. (LScE, Dumfries and Galloway Standard)

In only three cases a comma (2)–(3) precedes non-restrictive that:

(2) We're working with the licensing board and licensed trade associations to encourage late night catering so that people are eating whilst they are drinking and also with the Scottish Executive to pursue fire safety cigarettes, that are designed to go out fairly easily. (HIScE, Inverness Courier)

(3) The Moffat branch of the charity, that supports the serving and ex-service community and their families, is to have its first standard dedicated. (LScE, Dumfries and Galloway Standard)

4.2 Syntactic functions of the relativiser that

Relativisers (excluding zero, i.e. the omission of the relativiser in a relative clause) typically fill a subject gap in relative clauses. As figure 3 shows that typically functions as a subject in news writing. That is fairly common as an object and uncommon as an adverbial. No cases of that functioning as a complement were found in the present data (for classification of relativisers see Quirk et al. 1985: 1248–50).

Figure 3

Figure 3. Syntactic functions of that in HIScE, LScE, IrE and StE.

There is a clear correlation in the distribution of the syntactic functions of that between the ESSE varieties, and again between IrE and StE. That as a subject (4) is much more common in the two last mentioned in comparison to the ESSE texts (two-tailed chi-square with Yates’ correction: χ2= 5.521, df = 1, p-value = 0.0188). As an object (5) and adverbial (6) that is more frequent in the ESSE varieties in comparison to StE and IrE.

(4) She said: “We used to have a service that took three-and-half hours.” (LScE, John O'Groat Journal and Caithness Courier)

(5) “We will be honouring all the bookings that we have in the pipeline.” (HIScE, The Oban Times)

(6) “But, in practice, looking at the cases that are occurring in the Far East, they are people who are getting flu in the way that we normally get flu, through what we breathe.” (LScE, The Scotsman)

Prepositional complementation with that is accepted only with stranded prepositions, and this construction occurs when that functions as an object. Pied-piping, which means that the preposition precedes the relativiser, is not possible with that, but is the most typical means of prepositional complementation with wh-relatives.

Function PComp (N) PComp (%)

Table 2. The relativiser that as a prepositional complement in HIScE, LScE, IrE and StE.

Prepositional complementation with that is not frequent overall, but in HIScE nearly every tenth that relative clause is prepositional (7).

(7) "From our previous experience we expect that many of them will continue to deliver some or all of the food activities that they received funding for well into the future. (HIScE, Stornoway Gazette)

4.3 Gender type of the antecedent with the relativiser that

As mentioned earlier, in Older Scots that occurred with personal/human and nonpersonal/inanimate antecedents. After the introduction of personal relativisers into English, that came to be used predominantly as a nonpersonal relativiser. Biber et al. (1999) report that who is the most common relativiser with human antecedents in news writing. Other relativisers in this function are infrequent. Biber et al. (1999: 613) also remark that that is commonly found in speech with personal head nouns such as ‘people’, ‘women’ and ‘children’. This finding gains support, for example, from Tagliamonte et al. (2005: 91), who state that in a variety of Lowlands Scots spoken in Cumnock the relativiser that in subject function occurs much more frequently with human antecedents than zero and who. Also, human antecedents are more frequent than inanimate and collective antecedents with that in their data.

I have coded the antecedents in my data for three distinct types according to the Simplified Animacy Hierarchy introduced by Huber (2012): human NP, animate NP and inanimate NP. The class of animate NPs comprises collective nouns and animals. In the analysis of collective nouns I have followed the description provided by Quirk et al. (1985: 316–7), i.e. this category consists of nouns referring to groups of people such as army, organisation and majority.

Figure 4

Figure 4. Gender type of the antecedent with that in HIScE, LScE, IrE and StE.

As figure 4 shows that occurs predominantly with inanimate antecedents (10) in all the studied varieties. When we compare ESSE and IrE findings, we notice that IrE relativises only inanimate antecedents with that. Animate antecedents (11) are somewhat common in ESSE and StE. Human antecedents (12) with that are very rare, but they do occur. There are only ten cases across the data, majority of them are found in the HIScE material, and five of these instances concern the use of the lexical item ‘people’.

Inanimate antecedent:

(11) “I hope this is a trend that will continue for some years to come.” (LScE, Border Telegraph)


(12) “It is the council that is allowing people to get 2am licences in residential areas.” (HIScE, The Oban Times)

Human antecedent:

(13) “Of the people that have been named so far – Simon Hughes, Mark Oaten and Ming Campbell – I think that Ming is the one who has the most clear range of qualities we need in a leader, ” he stated. (HIScE, The Inverness Courier)

These findings show that the relativiser that is not the preferred choice in ESSE when relativising human and animate antecedents. This is one of the areas in the relativiser that use in which variation to spoken language is well reflected. While spoken Scots makes extensive use of that with human head nouns, news writing clearly follows StE norms relativising inanimate antecedents.

4.4 Definiteness of the antecedent: definite vs. indefinite

As a last feature, I will discuss the type of the antecedent from the point of view of definiteness. As figure 5 below shows the ESSE varieties function very similarly, and prefer indefinite antecedents (14) with that. Definite antecedents (15) are also common. The pattern is similar in IrE. StE seems to favour a completely different strategy and typically relativises definite antecedents with that.

(14) “We have to get a good system that works.” (HIScE, Stornoway Gazette)

(15) RPS will modify the existing gas control systems that are present on the site to direct the gas safely from the site to the generator before processing it for use. (LScE, Press and Journal)

Figure 5

Figure 5. Definite vs. indefinite antecedents with that in HIScE, LScE, IrE and StE.

Quirk et al. (1985: 1251) state that relativiser that as subject and zero relative as object are preferred to the relativiser which when the antecedent is an indefinite pronoun. Indefinite pronouns as antecedents are indeed found more commonly with that and the zero relative than with which in the current data. However, that provides mixed results across the data.

Variety Indef.pronoun (N) Indef.pronoun %
HIScE 16 13.9%
LScE 6 4.2%
IrE 4 12.5%
StE 2 3.2%

Table 3. The relativiser that with indefinite pronoun antecedents.

As table 3 shows, in HIScE and IrE indefinite pronoun antecedents (16) are found in more than every tenth that relative clause, while figures for LScE and StE are 4 per cent and less.

(16) He said that the purpose of the meeting was to review everything that had been done so far and to decide where to go from here. (HIScE, West Highland Free Press)

5. Conclusions and discussion

To conclude, the preliminary findings of my study indicate that the use of the relativiser that in written ESSE mainly complies with the rules of StE presented by grammarians. However, some traces of dialectal influence as well as regional differences are present. The relativiser that is common in news articles, and it is most frequent in LScE. This is expected, because earlier studies have shown that that is the predominant relativiser in spoken Scots. HIScE, on the other hand, has been described as a mainstream dialect of English having a close relationship to StE (Trudgill and Chambers 1991: 2–3). Therefore, it is not surprising that overall that occurs less frequently as a relativiser in HIScE in comparison to LScE.

Another feature that shows influence from spoken language is the frequent use of that with antecedents that are indefinite pronouns. This feature is best evidenced in HIScE. However, the most surprising finding in the current study is the frequency of non-restrictive that in StE news writing, in which it is found in almost every fourth that relative clause. Does this finding suggest that StE is gradually moving back towards the earlier use of that as a non-restrictive relativiser?

Although that is predominantly a restrictive relativiser in ESSE news writing, it also occurs in non-restrictive relative clauses. As mentioned above, non-restrictive that is an accepted feature of spoken Scots, and its use in ScE news may be due to dialectal influence from spoken language. On the other hand, we saw that non-restrictive that is common in StE news texts. Therefore it is possible that StE influences non-restrictive that usage in ScE rather than spoken Scots. Because non-restrictive that is such an under-researched subject it would be fruitful to include it in future studies concerning relativisation, so that we would find more about its functions in English and see if its use is in fact in increase.

Huber (in preparation) suggests that news writing as a genre is less formal than, for example, administrative and academic writing. The rather common use of non-restrictive that in the StE data lends support to this view, but because the results are not uniform across the material under investigation no definite conclusions can be made. Also of interest is the fact that the ESSE texts are taken from online news texts, which, on the other hand, could be expected to be less formal than printed news articles. However, majority of non-restrictive that in the StE data and all IrE instances are from printed newspapers. This result therefore suggests that online news writing is not less formal than printed news writing.

The findings of this study show that grammatical variation is found between regional standards of English. Studies on English dialects have shown the great amount of variety that exists in speech, but variation in written language and in the form that could be defined as ‘educated’ or ‘standard’ has not been addressed as vigorously. This is yet another line of interesting investigation for future.


[1] In Aitken 1984b, this term refers to the dialect spoken by the most educated group of Scottish English speakers. In this paper, the term is used to refer to the degree of formality of language. News writing as a genre represents educated language, because journalists are typically well-educated in writing and news writing is edited.

[2] According to Robinson (2005: xiii), the Older Scots period extends until 1700. It comprises the periods of Early Scots (–1450) and Middle Scots (1450–1700).

[3] RRs = restrictive relative clauses

[4] Huddleston and Pullum (2002) use terminology that differs from the traditional terminology describing relative clauses: supplementary relative clauses correspond to non-restrictive relative clauses including sentential relative clauses and integrated relative clauses to restrictive relative clauses.


Aitken, A.J. 1984a. “Scottish accents and dialects”. Language in the British Isles. ed. by Peter Trudgill, 94–114. Cambridge: Cambridge University Press.

Aitken, A.J. 1984b. “Scots and English in Scotland”. Language in the British Isles. ed. by Peter Trudgill, 517–532. Cambridge: Cambridge University Press.

Ball, C. N. 1996. “A diachronic study of relative markers in spoken and written English”. Language Variation and Change 8: 227–258.

Beal, J. 1997. “Syntax and Morphology”. The Edinburgh History of the Scots Language. ed. by C. Jones, 335–377. Edinburgh: Edinburgh University Press.

Beal, J.C. & K.P. Corrigan. 2002. “Relatives in Tyneside and Northumbrian English”. Relativisation on the North Sea Littoral ed. by P. Poussa. München: Lincom Europa 2002.

Bergs, A. 2001. Modern Scots. München: Lincom Europa.

Biber, D., S. Johansson, G. Leech, S. Conrad & E. Finegan. 1999. Longman Grammar of Spoken and Written English. London: Longman.

Caldwell, S.J.G. 1974. The Relative Pronoun in Early Scots. (=Mémoires de la Société Néophilologique de Helsinki, 42). Helsinki: Société Néophilologique

Douglas, F. 2009. Scottish newspapers, language and identity. Edinburgh: Edinburgh University Press.

Görlach, M. 2002. A Textual History of Scots. Heidelberg: Winter.

Herrmann, T. 2003. Relative clauses in dialects of English: a typological approach. PhD dissertation, University of Freiburg. Available at, accessed Feb 2, 2012.

Herrmann, T. 2005. “Relative clauses in English dialects of the British Isles”. A Comparative Grammar of British English Dialects, ed. by B. Kortmann, T. Herrmann, L. Pietsch & S. Wagner, 21–123. Berlin & New York: Mouton de Gruyter.

Hillberg, S. (forthcoming). “Development of the relativization system in Scots and Scottish English”. Proceedings of the 24th Scandinavian Conference of Linguistics. Joensuu.

Huber, M. 2012. “Syntactic and variational complexity in British and Ghanaian English. Relative clause formation in the written parts of the International Corpus of English”. Linguistic Complexity: Second language acquisition, indigenization, contact, ed. by B. Kortmann & B. Szmrecsanyi, 218–242. Berlin: Walter de Gruyter.

Huber, M. (in preparation). “Morphosyntactic and stylistic complexity in British and Ghanaian English: a comparison of relative clauses in the written parts of the International Corpus of English”.

Huddleston, R. & G. K. Pullum. 2002. The Cambridge grammar of the English language. Cambridge: Cambridge University Press.

King, A. 1997. “The Inflectional Morphology of Older Scots”. The Edinburgh History of the Scots Language. ed. by C. Jones, 156–181. Edinburgh: Edinburgh University Press.

Kortmann, B. & K. Lunkenheimer, eds. 2011. The electronic World Atlas of Varieties of English [eWAVE]. Leipzig: Max Planck Institute for Evolutionary Anthropology. 21 Jan. 2012.

Macafee, C. 1983. Glasgow. Amsterdam & Philadelphia: John Benjamins.

Macafee, C. 1996. The case for Scots in the 2001 Census. 26 Apr. 2012. ttp://

Macafee, C. 2011. Characteristics of non-standard grammar in Scotland. 26 Apr. 2012.

Macaulay, R. K. S. 1991. Locating dialect in discourse: the language of honest men and bonnie lasses in Ayr. New York: Oxford University Press.

McClure, J.D. 1994. “English in Scotland”. The Cambridge History of the English Language. ed. by Robert Burchfield, vol. V, 23–93. Cambridge: Cambridge University Press.

Meurman-Solin, A. 2000. “Geographical, socio-spatial and systemic distance in the spread of the relative who in Scots”. Generative Theory and Corpus Studies: a dialogue from 10 ICEHL ed. by R. Bermúdez-Otero, D. Denison, R. M. Hogg & C. B. McCully, 417–38. Berlin: Mouton.

Millar, R. M. 2007. Northern and Insular Scots. Edinburgh: Edinburgh University Press.

Miller, J. 1993. “The Grammar of Scottish English”. Real English: The Grammar of English Dialects in the British Isles. ed. by J. Milroy & L. Milroy, 99–138. London: Longman.

Miller, J. 2008. “Scottish English: morphology and syntax”. Varieties of English. 1, The British Isles, ed. by B. Kortmann & C. Upton, 299–327. Berlin: Mouton.

Moessner, L. 1997. “The Syntax of Older Scots”. The Edinburgh History of the Scots Language. ed. by C. Jones, 112–155. Edinburgh: Edinburgh University Press.

Petyt, K. M. 1980. The study of dialect: An introduction to dialectology. London: André Deutsch.

Quirk, R., S. Greenbaum, G. Leech and J. Svartvik. 1985. A Comprehensive Grammar of the English Language. New York: Longman.

Robinson, M., ed. 2005 [1985]. The Concise Scots Dictionary. Edinburgh: Edinburgh University Press.

Romaine, S. 1980. “The relative clause marker in Scots English: diffusion, complexity and style as dimensions of syntactic change.” Language in Society 9: 221–49.

Romaine, S. 1982. Socio-historical Linguistics: Its Status and Methodology. Cambridge: Cambridge University Press.

Romaine, S. 1985. “Syntactic variation and the acquisition of strategies of relativization in the language of Edinburgh schoolchildren”. Papers from the Third Scandinavian Symposium on Syntactic Variation ed. by Sven Jacobson, 19–33. Stockholm: Almqvist & Wiksell.

Rydén, M. 1979. An Introduction to the Historical Study of English Syntax. Uppsala: Almqvist & Wiksell.

Shuken, C. 1984. “Highland and Island English”. Language in the British Isles ed. by Peter Trudgill. Cambridge: Cambridge University Press.

Smith, J.J. 2000. “Scots”. Languages in Britain and Ireland. ed. by G. Price, 159–170. Oxford: Blackwell.

Tagliamonte, S., J. Smith & H. Lawrence. 2005. “No taming the vernacular! Insights from the relatives in northern Britain”. Language Variation and Change 17: 75–112.

Tottie, G. 1997. “Relatively speaking: relative marker usage in the British National Corpus”. To explain the present: Studies in the Changing English Language in honour of Matti Rissanen ed. by Terttu Nevalainen & Leena Kahlas-Tarkka, 465–481. Helsinki: Société Néophilologique.

Trudgill, P. and J.K. Chambers. 1991. “Introduction: English dialect grammar”. Dialects of English: Studies in Grammatical Variation ed. by Peter Trudgill & J.K. Chambers. New York: Longman.