'I'll cross dat brig whin I come til him': grammatical gender in the Orkney and Shetland dialects of Scots.
The purpose of the current paper is to investigate possible remnants of a grammatical gender system found in the Orkney and Shetland dialects of Scots. The fact that Orkney and Shetland dialects have something of a grammatical gender system has long been established. Its origin, however, is disputed. Robert McColl Millar (2007: 67) remarks that
In Shetland dialect, a number of nouns can be referred to using sex-specific pronouns, so that sun can be referred to as he, while kirk, 'church', can be referred to as sho, 'she'. Whether this is a relict of the Old Norse grammatical gender system, or is an extension of the habit found among all English speakers to personify certain objects considered affectionately or used regularly, referring to them as she, is beyond our scope [...].
Similar uses of the pronouns he and she can also be found in Orkney dialect. This usage does not, however, amount to a full grammatical gender system visible in nouns and adjectives. Hugh Marwick (1929: xxix) and Jakob Jakobsen (1928-32: xxxvii) both find, in their respective studies of Nom substrate in Orkney and Shetland dialect, that the Old Norse inflectional system is mostly lost in the nineteenth and twentieth century dialects they were studying. However, both observe that Old Norse morphology survives in fossilised form in certain nouns (Marwick 1929: xxix, Jakobsen 1928-32: xxxvii-xli). For example, the strong masculine nominative--r is fossilised in blouster ('strong gale', ON blastr); the weak masculine nominative--i in flackie ('straw mat', ON flaki); weak masculine accusative -a ending as--o in klavo ('strap for horse collar', ON klafi); and the feminine nominative--a as--o in berg-gilto ('the wrasse', ON gylta), along with numerous other examples listed by Marwick and Jakobsen. The question is whether it is possible to detect any remnants of a grammatical gender system in the Orkney dialect apart from this type of fossilisation. Marwick (1929: xxx) notes:
The old differentiation of gender in the Orkney Nom is extremely difficult to discover to-day, when the former habits are so overlaid by Scottish usage. In a few connexions, however, the older genders still persist.
Of those 'few connexions,' Marwick (1929: xxx) mentions that weather and time phenomena are spoken of in the masculine, while fish are generally referred to as feminine. He then adds that '[i]n general, concrete objects, e.g. gun, nail, & c., are spoken of as feminine, and that is very common even yet' (Marwick 1929: xxx).
The purpose of this paper is to address some of the questions that Millar (2007) does not have the scope to address. Firstly, it seeks to address the question of whether the Orkney or Shetland system shows any sign of being a relict of Old Norse grammar, which has a fully developed gender morphology in nouns and adjectives. The historical background to this question is the process of language shift in the Northern Isles, in which Scots superseded the Old Norse dialect Norn. This process took several hundred years and cannot be dealt with here (but see Barnes 1998; Knooihuizen 2005; Millar 2008; Ljosland 2012; Millar 2012). However, in order to address the issue fully it is also necessary to consider gender in other forms of (relatively) modern English, as well as in Old English and Gaelic as Nom's geographical neighbours.
This paper investigates not only the workings of a grammatical gender system in the Orkney and Shetland dialects of Scots, but also addresses the question of its origin by comparison to other historical and contemporary varieties of English and Scandinavian. Also, influence from Gaelic, which employs two genders, cannot be excluded for the reason of its geographical proximity to the Earldom of Orkney and because of the possibility that Middle Scots could have carried Gaelic influence before it reached Orkney and Shetland. No attempt will be made in this paper to explain 'why' a certain noun has a certain gender, as Wales (1996: 138) does when she, supposedly in jest, says that 'On the surface it is hard for the non-Orkney speaker to understand why a door should be referred to as sho ('she'), or a ball by he; except that a ball could be seen as a 'dynamic' entity, and the door [...] as stubbornly resistant!'
The initial hypotheses for this paper will therefore be:
Hypothesis 0: There is no non-standard usage of gendered pronouns in Orkney and Shetland.
Hypothesis 1: (A) system(s) of gendered pronouns found in Orkney and Shetland Scots retains features carried over from the grammatical gender morphology of Norn.
Hypothesis 2: (A) system(s) of gendered pronouns found in Orkney and Shetland Scots retains features from the gender system of Old English.
Hypothesis 3: (A) system(s) of gendered pronouns found in Orkney and Shetland Scots contains features from the gender system of Gaelic.
Hypothesis 4: (A) system(s) of gendered pronouns found in Orkney and Shetland Scots has developed independently of Nom/Old Norse, Scots/Old English and Gaelic and is today (a) system(s) exclusive to the Northern Isles.
Hypothesis 5: (A) system(s) of gendered pronouns found in Orkney and Shetland Scots is similar to system(s) found in other dialectal (non-standard) varieties of modern English.
3. GENDER IN ENGLISH
Two recent studies deal at length with gender in relatively present-day British English: Wagner (2005) and Wales (1996), and of these two Wagner takes the most systematic approach and will therefore chiefly be referred to here. Also, Hernandez (2011: 73-94) includes a section on gendered pronouns in English, while Untersbeck and Rissanen (eds. 2000) provide a broad scope of gender in languages across the world.
Whereas a full grammatical gender system, as for instance in German, Norwegian and Old English, requires agreement between nouns, adjectives and pronouns, 'gender' in modern English has only been described with reference to gendered pronouns, and that is also the scope of this paper.
There are two major systems for gender assignment worldwide, and these may be termed semantic versus formal (Wagner 2005: 222). Semantic systems typically distinguish between animate and inanimate nouns, alternatively human and non-human, where the animate or human class is assigned masculine or feminine gender. In formal systems, on the other hand, gender assignment does not rely on semantics but on such things as morphological or phonological criteria (ibid.). However, many languages employ a mixture of these two systems. In the case of English, '[i]t is generally known that English inherited a formal gender system from its Germanic parent language which, roughly between the tenth and fourteenth centuries, was gradually replaced by the semantic 'natural' or 'logical' gender system that is known from Pr[esent] D[ay] E[nglish], the change having been completed in early Middle English' (Wagner 2005: 225).
However, when examining corpora of various forms of modern English, it becomes apparent that gendered pronouns have a certain currency. Wagner (2005) reviews studies spanning spoken and written varieties of English including American and Canadian novels, spoken standard American and Canadian English, and Tasmanian vernacular English. The conclusion of this review is that when gendered pronouns are employed, 'she is the gendered pronoun of choice in non-dialectal English. [...] In everyday, casual spoken English, possibly world-wide, the pronoun of choice when referring to an inanimate noun and wishing to add extra information is [...] a form of she. Mostly, this 'extra' has been identified as some sort of emotional information, either positive or negative. In contrast, the pronoun signifying non-involvement or simply disinterest is it' (Wagner 2005:274). Wagner's own study is concerned with the dialects of south-western England and Newfoundland (brought from south-western England), and she finds that the system for the 'spoken Standard' where the pronoun she prevails stands in sharp contrast to the dialectal system of her geographical areas, where 'he and the corresponding object form(s) occur in a large number of slots that are [otherwise in the spoken standard] occupied by she [...]' (Wagner 2005: 275). In 'traditional 'West Country' (Southwest) dialects and (West-Country based) Newfoundland English [...] neuter pronouns are traditionally only employed for mass nouns, while count nouns trigger masculine forms' (Wagner 2003, abstract: 1). Wagner (2005: 348-49) sums up the gender assignment systems found in written Standard English, as well as what she calls the 'spoken standard' and the traditional West Country dialects of England by presenting the following tables:
(Written) Standard English (as presented in grammars): Humans Masculine or feminine (according to sex) Animals Supposedly neuter, but also masculine and feminine (not always according to sex) Count nouns Boats, countries, etc. can also be feminine ([right arrow] metaphorical gender) Mass nouns Neuter; if anything else, then feminine abstract nouns Neuter; regionally also feminine ([right arrow] AusE). Spoken Standard Humans Masculine or feminine (according to sex) Animals Generally not neuter, but masculine and feminine (not always according to sex) Count nouns Can be masculine and feminine, although feminine forms seem to be more frequent (cf. studies by Svartengren, Mathiot/Roberts and also Pawley) Mass nouns Can be feminine, masculine and neuter, but are predominantly neuter Abstract nouns Although generally neuter, often feminine in special constructions ([right arrow] non-referential she etc.) Traditional West Country dialects (19th century) Humans Masculine or feminine (according to sex) Animals Generally masculine (often even when female [right arrow] cows in SED). Count nouns Masculine Mass nouns Neuter Abstract nouns Neuter
Using data from a subset of the Freiburg English Dialect Corpus, spanning the whole of England, Hernandez (2011:93) confirms the dominance of the masculine over the feminine pronoun in dialectal speech. Hernandez (2011: 85) further develops the idea of the likelihood of gendered pronouns occurring along an animacy hierarchy going from human animate > non-human animate > inanimate count > inanimate mass nouns, with decreasing degrees of individuation. Hernandez finds that in dialectal speech gendered pronouns occur along the entire continuum, declining to the right, while the neuter pronoun declines to the left. However, whereas masculine pronouns span the whole hierarchy with 2.5% gendered pronouns in the inanimate count category and 0.3% in the inanimate mass category, feminine pronouns account for 1.5% of the inanimate count category and do not occur at all in the inanimate mass category (Hernandez 2011: 85). In terms of geographical distribution, gendered pronouns are most common in the South-West and South-East of England, declining in the Midlands and least common in the North (Hernandez 2011: 92). However, even in the areas where gendered pronouns occur most frequently, they account for a maximum of 7% of the tokens in the inanimate count category. Sadly, Hernandez does not include data for Scotland.
4. DATA MATERIAL
In the absence of a digital corpus of spoken Orkney and Shetland dialect, the current paper draws its data from prose fiction written in these dialects. There is a time span of a century in the publication dates of the texts from 1880 to 1987. The Orkney dialect texts are Walter Traill Dennison's story 'Why the Hoose o' Hellsness was Brunt', from his collection The Orcadian sketch-book: being traits of old Orkney life written partly in the Orkney dialect (1880), and Christina Costie's short story collection The Collected Orkney Dialect Tales of C. M. Costie, published in 1976 but written in the 1940s, 50s and 60s. The Shetland texts examined are James Stout Angus: Eels (poem, 1877); Basil Ramsay Anderson: Auld Maunsie's Cru (poem, 1888); three short stories 'A Wrastle Wi' a Hen', 'Lowrie Buys a Ford', and 'Lowrie At Da Exhibishon' from the collection Lowrie by Joseph Gray (1991 edition, but first published 1933) and John Graham: Shadowed Valley (novel, with Shetland dialect dialogue, 1987).
In the following, referent nouns will be categorised as count nouns, mass nouns or abstracts in accordance with Wagner (2005). Human referents are excluded, while animal and body part referents are dealt with separately. Firstly, however, I would like to comment more specifically on certain semantic categories, namely ships, boats and vehicles which seem to constitute a special referent class in most forms of English. Thereafter, I would like to comment on some referent classes which are of particular interest to the question of a Norn origin, for reasons which will become apparent. These referent classes are weather phenomena, tide and time, seasons and the sun and moon.
Ships, boats and vehicles
For the purpose of this investigation, ships and boats are excluded for the reason that these belong to a special referent class and are commonly referred to using the feminine pronoun also in Standard English (Wagner 2005: 231, Hernandez 2011:78). Examples from my text material can be seen in 0a and 0b:
0a: An' in sheu cam' straicht tae the Bay o' Kettletaft, an' anchored there. (Orkney, Walter Traill Dennison)
0b: Lay her aboot! (Shetland, John Graham)
In general present-day English, the category of ships offers a two-way choice between personal 'she' and non-personal 'it' (Wagner 2005: 231), and this usage can extend to other vehicles as well (Wagner 2005: 232), and there is even some evidence to suggest that large machinery can fall into this category. A good example of this usage in Shetland dialect can be seen in 'Lowrie buys a Ford', which contains 38 tokens of she/her=car, in both Shetland dialect-e.g. 'Bit why did doo no tak her hame?'--and in Standard English 'Power, why she will easily take you up the Stoney Hill on top'. The story also has 2 tokens of road roller=she, as well as 3 tokens of she/her=engine: 'Joannie says, I tink shu's missin'. What, says I? (fur I kent we hed nae wuman passenger ta faa oor ower). It's da engine, man. Na, heth, says I, shu's here yit, I can see pairt o' her trow da sems.'
While Wagner (2005: 259) observes that in studies of more standard varieties of English, whether phenomena can be spoken of as 'she' as part of a category she terms 'non-referential she' Hugh Marwick (1929: xxx) observes that in Orkney weather phenomena are usually spoken of as masculine. Millar (2007: 68-69) also notes that an older edition of Robertson and Graham's Shetland grammar lists the masculine gender for weather, giving the example 'He's a cowld day' (despite 'he' seemingly referring to 'day' here). The observation about masculine gender for weather is to a large extent confirmed by my data. For example, in Orkney dialect:
1a: Noo he was a sooth-aesterly gale (Costie 1976)
1b: Bae the time sheu won heem he wis blawan for a' he wis wort (Costie 1976)
1c: He was piercan cowld outside (Costie 1976)
1d: He was a closs miffy night, an' the peur body couldna get a breath o' air (Costie 1976)
And of the darkness of the night:
1e: He wis as dark as you been under a pot (Costie 1976)
The Shetland texts also contain some scattered tokens of weather=he.
1f, a particularly calm passage: ... an' fur aa da wadder he wis, we cood as weel bune apo Eelawatter. (Gray 1933)
1g, gale: Da wind got up trowe da nicht and be da moamin he was a flying gale. (Graham 1987)
1h, temperature: We wima dat lang i da water but lang enyoch whin he's freezin you tae da bane. (Graham 1987)
And one token of a 'moorie' (blizzard) being referenced by 'it':
1i, 'moorie': He set aff yesterday momin for Stromfirt juist afore da moorie cam. It wid a come apon him I da hill. (Graham 1987)
The form 'it' rather than 'hit' here suggests influence from Standard English.
Tide and time
Marwick (1929: xxx) also observes that time and tide are referred to as masculine, and gives the following examples:
2a: 'What time 'o day is he?'; 'He'll be gettan on tae 9 o'clock.' (Time)
2b: 'He was low water then'; 'He was beginnan tae flow.' (Tide)
Marwick's observation is supported in my data, as can be seen in 2c-2e:
2c: He wus i' the mirk o' the mornin' (Dennison 1880)
2d: He was a peerie while afore the king's men got a boat (Dennison 1880)
2e: Bae this time he wis wearan on by eleven o'clock (Costie 1976)
However, Dennison's story Why the Hoose o Hellsness was Brunt opens with a contradiction regarding the gender of 'year': (2f) 'Aye, bairns, he's jeust fower scor' an' fifteen year sin' de Forty 'five' is followed by (2g) 'Sheu wus a sair ga'n year amang the gentry.' Either this contradiction shows an uncertainty on the part of the author of what the gender of 'year' should be or, since the first 'year' is plural while the second is singular, 'he' in 2f could refer to a general sense of 'time' rather than a specific 'year', which would resolve the contradiction.
Supporting Hugh Marwick's observation (above), the tide and its ebb and flow is also referred to as masculine, as can be seen in 2 h-j:
2h: He wus low water whin Ayrie geed i' the geo. (Dennison 1880)
2i: He'll be a peerie while yet afore he's flowan watter. (Costie 1976)
2j: He was just begood tae ebb. (Costie 1976)
The use of gendered pronouns 'especially he, for weather, time, and other natural phenomena' is by the Oxford Companion to the English Language (McArthur 1992:731) regarded as 'probably of Norn origin'. This will be discussed further below.
The names of the four seasons in the Orkney and Shetland dialects are voar (also 'spring', under influence from Standard English), simmer, hairst and winter. These are referred to as masculine:
3a: He was a bonnie spring day (Costie 1976)
3b: He is simmer an hardly dark ava (Costie 1976)
3c: He wis a bad hairst (Costie 1976)
Seeing the masculine here is not surprising, as it is in accordance with other time expressions and weather/atmospherical conditions. In Old Norse, the gender of var (spring) and vetr (winter) is also masculine, while sumar (summer) is neuter. However, snmar/sommer in modern Norwegian is masculine. Old Norse haust (autumn) is also neuter, however the Oxford English Dictionary has it as originally masculine, and in Modern Norwegian haust/host is also masculine. In Old English, spring/spryng, sumor, hoerfest/herfest and winter are all masculine. In this light, it is not surprising to find the four seasons referred to as masculine in the dialect material.
However, the use of the masculine pronoun in reference to weather conditions, tide, time and seasons is surprising when viewed against finds from studies of other varieties of present-day English. In Canadian English it was found that '[w]hile masculine reference to any type of inanimate denotatum is extremely rare, no examples at all were found in which a native [Canadian] English speaker used he to represent an intangible, difficult-to-identify type of denotarum.' (Morris 1991: 164, quoted from Wagner 2005: 272). Svartengren 1928, who used a corpus of contemporary mostly American fiction, also found gendered pronouns to be extremely rare, but found that '[n]ature and natural objects not worked upon by man' and the sub-categories '[s]easons, periods' and 'fire, temperature, weather conditions, ice, snow' would sometimes trigger a feminine pronoun (quoted from Wagner 2005: 268). The feminine pronoun was also found in this type of context by Wagner (2005: 259), who includes weather in what she terms the 'non-referential she'.
The sun and moon
In order to examine the hypothesis of a Norn origin, an interesting indicator would be the gender of the sun and the moon. Marwick (1929) makes this point:
While in modern English usage the moon is often personified as a female, in the old Orkney tongue moon was masculine [...] (Marwick 1929: xxx).
The modern English poetic usage, personifying the sun and moon, has adopted the Latin or Romance (largely French) gender for sol (masculine) and luna (feminine) instead of retaining the Germanic grammatical genders where the sun is feminine and the moon masculine. If the Orkney and Shetland dialects had inherited these genders from Nom, one would expect them to follow the Old Norse genders: sol (fern.) and mani (masc.). Interestingly, Robertson and Graham (1991: 2), in their Grammar and Usage of the Shetland Dialect which draws on a number of nineteenth century dialect texts, has the moon listed as one of 'a smaller number' of words which are feminine. An example may be found in Basil Ramsay Anderson's poem 'Auld Maunsie's Cru' (1888):
4a: An', bricht an' bonnie, raise da mune, Auld Elder Rasmie o' da mill/Grew restless as shu neared da hill.
The Orkney and Shetland dialects therefore have different genders for the moon, which means that if the fact that 'moon' is masculine in Orkney dialect is to be seen as a relict of Old Norse, it implies that the gender of the moon must have changed in Shetland dialect at some later point in time, possibly under influence of poetic usage in Standard English.
What about the sun? While Marwick (1929) does not comment on the gender of the sun in Orkney dialect, Robertson and Graham (1991) explicitly say that the sun is masculine in Shetland dialect. Shetland dialect therefore seems to follow the general modern English convention of referring to the moon as feminine and the sun as masculine, while Orkney dialect differs in referring to the moon as masculine. I have not succeeded in finding any pronominal references to the sun in the text material studied for this paper.
The case for Orkney dialect having retained Old Norse genders for the sun and the moon must therefore be regarded as indecisive at best. While the occurrence of the masculine pronoun in reference to the moon certainly is at odds with modern Standard English, it could be the result of a more general tendency to refer to weather, tide and time phenomena as masculine rather than reflecting a memory of the Old Norse gender assignment.
Interestingly, Wagner (2005: 307-8) in her investigation of traditional dialects of South-West England also finds that the masculine pronoun is used with reference to the moon there. While not committing to any one explanation, she suggests two reasons for this; one being that the referent has been personified as 'the man in the moon' and the other that it retains the Germanic grammatical gender.
When referring to human beings, biological sex often takes precedence over grammatical gender even in languages that have full gender morphology, such as in modern Norwegian where barn 'child' is grammatically neuter but usage shows that he and she are instead often applied according to the child's sex (see also Dahl 2000: 105-106). With animal referents, the dilemma arises of whether to refer to the animal as it or a gendered pronoun, and furthermore whether the gendered pronoun must agree with the animal's sex or not. There is a case for arguing that animals and fish may constitute a special referent class which, like ships and vehicles, is more likely to trigger a gendered pronoun in many varieties of spoken English: for example, gendered pronoun references to animals were found to far outnumber neuter it in Wagner's survey of south-western English dialects (Wagner 2005: 254). Dahl (2000: 100-101) suggests that the semantic and the formal principles of assigning gender may cut off at various positions in the animacy hierarchy, so that some languages have the cut-off-point within the 'non-human animate' category, treating 'higher' animals as 'persons' and therefore applying the semantic principle, while 'lower' animals are categorised with inanimate nouns and applied to the formal principle of gender assignment. This would perhaps explain why the likelihood of a gendered pronoun occurring was found by Wagner (2005: 258-9) to rise with the speaker's personal involvement with animals, so that a farmer, hunter or fisherman is more likely to use gendered pronouns than people in occupations or environments that do not include animals. This is relevant in traditional farming and fishing communities such as Orkney and Shetland. Some examples would be:
5a, cow: I was tinkin ta keep wir young milkin coo an tak her wi wis (Shetland, Graham 1987)
5b, hen: Man, tak her attween your haands, bit noteece it [= that] shu's aye lookin i' your face (Shetland, Gray 1933, + 12 further tokens of she/ her=hen in the same story)
5c, cat: ... shu raise wi' a yall laek a steemer (Shetland, Gray 1933)
In the grey area of the cut-off zone between semantic and formal gender assignment, gender conflicts arise, and bring up the question of whether the pronoun is chosen with reference to the word or the actual animal (Dahl 2000: 111). Therefore, gendered references to animals are not always consistent, even in consecutive references to the same animal, as in the case here with an orphaned lamb:
5d, lamb: I sail fin da bottle ta gie him a sook (Shetland, Graham 1987)
5e, lamb: We dunna hae da milk ta gie h as a caddy (Shetland, Graham 1987)
Syntactic function is also relevant: In a study of Canadian English, a gendered pronoun was typically preferred if the animal was foregrounded as the topic of conversation (Morris 1991, reported in Wagner 2005: 271).
Most interestingly, the pronoun does not necessarily have to reflect the animal's biological sex. In a syntactically foregrounded position, he was found to function as a neutral pronoun and she as specifically female (Morris 1991, reported in Wagner 2005: 271). Examples from the text material under scrutiny here could perhaps be:
5f, polar bear: Da polar bear is a strange animal. You can never tell what wye he's goin ta behave. (Shetland, Graham 1987)
5g, polar bear: I seun sensed he wis failin, an' when thir wisna muckle more tullye left i' 'im, I got haad o' me knife an' stuck 'im right i' the hert. (Orkney, Costie 1976)
5h: A conger eel? Na heth it, boy; I ken 'm by his shape (Shetland, Angus 1877)
It seems relevant here that the biological sex of the polar bear or conger eel might not be known. Otherwise, both Marwick (1929) and Robertson and Graham (1991) list fish as something which is usually feminine, although I do not know whether an eel would fall in this category. A whale is also masculine, as demonstrated here in Shetland dialect:
5i: If he's below da flude-mark da laird canna claim him (Graham 1987)
And a whale's head in Orkney dialect:
5j: Min, if thoo're that silted wae could gae thee a whal'heid an a Greullie-Bjelkie tae boil 'im in. (Costie 1976)
This brings us on to the question of whether human or animal body parts
should be regarded as a special referent group. Wagner (2003: 234) finds pronominal references to body parts rare in her south-western English material, but notes a total of 11 masculine and no feminine references. Pronominal referents to body parts in Costie (1976) are of different genders: whale's head (masculine), lug ('ear', feminine), and hand (neuter). The two instances of body parts in the Shetland material are both masculine: Foot and head, both from Graham 1987.
Inanimate count nouns
Regarding inanimate nouns, the system which Wagner describes for the traditional dialects of South-Western England is characterised by a count-mass distinction, where gendered pronouns are common only with reference to count nouns. Masculine dominates this category. Hernandez (2011: 91-92) shows that the use of gendered pronouns with reference to inanimate count nouns is not exclusive to the South-West, but also exists in South-East England with a more balanced gender distribution, and less commonly in the Midlands where again masculine dominates. Gendered pronouns in this category are almost non-existent in the North of England (2011: 91-92). Hernandez (2011: 91-92) describes the distribution as a South-North continuum, although it must be said that even in the areas where gendered pronouns are most frequent they account for a maximum of 7% of the tokens.
Count nouns in Orkney dialect
Contrary to the distribution of masculine versus feminine gendered pronouns in England described above, Hugh Marwick (1929: xxx) observes that in Orkney dialect feminine pronouns dominate for count nouns: '[i]n general, concrete objects, e.g. gun, nail, &c., are spoken of as feminine, and that is very common even yet.' My study suggests that, generally speaking, this is indeed the case. However, there is also a small number of masculine referents, which will be discussed in the following.
The count noun referents of feminine pronouns in the Orkney material range from writin 'a document' to guinea, 'gun', ball 'bullet', silk (a specific silk garment), knife, table, coat, a piece of peat, a glass, egg, cubbie 'straw basket' and cap. Some sentences extracted from the Orkney text material are given below:
6a, document: Sheu wus a writin' o' muckle stead; for sheu wus no' paper, bit skin (Dennison 1880, + 3 more tokens of this document = she/her).
6b, guinea: [S]heu fand de guinea i' the hoi' o' the wa' whar' sheu left her (Dennison 1880)
6c, silk: Sheu wus a maist beautifu' strae coloured silk wi gowd floo'ers (Dennison 1880)
There are also examples of the feminine pronoun used to refer to concrete phenomena which are not objects, as such, but nonetheless count nouns, such as a salt water pool and a geo or coastal ravine:
6d, saltwater pool: Ollie Meur ran i' a sa'twater pow an' steud i' her a hoor ap tae the neck. (Dennison 1880)
6e, 'geo' ravine: The geo wus heech eneuch, bit her sides wur a' smeuth (Dennison 1880)
With the exception of the whale's head (see animal referents above), I have only found one example in the Orkney text material of a masculine pronoun being used for an inanimate count noun. The referent in this case is a stone:
6f, stone: An' abeun de yett wus a bonnie square free-steen wi' letters cuttid on him 'at nee bodie could read (Dennison 1880).
Neuter count nouns, on the other hand, are not quite as scarce in the Orkney text material. Referents are for example nail, hide, powder horn, hood, staff and hand:
7a, nail: [H]e saw lyan' at his fit, a muckle carval nail [...]. Seurly a kind providence pat hid there. (Dennison 1880)
7b, horn: Ayrie's pooder horn wus i' the pooch on his ither side; bit he teuk hid oot an' stappid hid i' his breest pooch (Dennison 1880)
The nail in 7a is neuter despite Marwick's statement that '[i]n general, concrete objects, e.g. gun, nail, &c., are spoken of as feminine [...]' (Marwick 1929: xxx). It is also interesting that the document which triggered four feminine pronominal references above, also appears with a neuter pronoun, albeit as a reference to the noun hide rather than writin. I believe the general modern English system now competes with the traditional Orkney system, the result being that inanimate count nouns can trigger the neuter 'it' as well as a gendered pronoun.
Count nouns in Shetland dialect
Perhaps the biggest surprise in the dialect text material is that Orkney and Shetland dialect differ greatly in their preference for which gendered pronoun to apply in reference to count nouns. Whereas in Orkney the preferred gendered pronoun is the feminine, count nouns in Shetland dialect tend to be referred to as masculine. Masculine referents here range from chair to hook, bonfire (3 tokens), fiddle, house, pig (a certain type of bottle, 2 tokens), kist 'chest', sail, brig 'bridge', sou 'wester, horn, barrow (4 tokens), cravat, telegraph pole, an iron, a burn 'stream' and interestingly also kirk 'church'. Some examples include:
8a, chair: I hae ta lead da coo an cerry Granny's shair. Come an gie me a haand wi him. (Graham 1987)
8b, house: Wir been here I dis hoose for near on fifty year noo--an I'm wae ta laeve him. (Graham 1987)
8c, bridge: I'll cross dat brig whin I come til him (Graham 1987)
8d, church: Was der mony at da Kirk? Yae, as full as I'm seen him for a braa start noo. (Graham 1987)
The dominance of the masculine pronoun for count nouns in Shetland dialect is confirmed in T. A. Robertson and John Graham's Grammar and Usage of the Shetland Dialect (1991: 2):
Common Nouns are either Masculine, Feminine or Neuter. Nouns which are Neuter in English are often Masculine or Feminine. The following are among those usually considered to be Masculine: aer, steid, schair, spade, sun. [...] A smaller number are usually Feminine, including: lamp, fish, kirk, mon, wirld. [...]
It is interesting, however, to note that in 8d kirk is masculine, while Robertson and Graham specifically list the word as feminine (especially since the two texts are by the same author!). In fact, in the data I have not been able to find any examples of count nouns being referred to by the feminine pronoun in the Shetland texts except the moon, ships, boats, cars and other vehicles, machinery, and human and animal referents.
Like in the Orkney material, however, some tokens of the neuter pronoun can be found, both in its standard form 'it' and Shetland 'hit'. Referents include bed, handle, crown bevel, car, tow 'rope', knife (2 tokens). Examples include
9a, bed: ... we sail tak it hame wi wis (Gray 1933)
9b, handle: So I gae hit a shiv frae me, bit I wiss I hed left hit alane (Gray 1933)
There are also three tokens in Gray 1933 of hit being used in reference to what the reader knows to be a vacuum cleaner, but for which the protagonist has no word other than describing it as a thing and a wonder.
As in the Orkney material, there seems to be some competition between the traditional gendered pronoun and the neuter, resulting in mixed references such as in 9c below:
9c, tea-kettle: He grips a pritty sheenin'-laek tay-kettle, an' poors a coarn o' watter an' sets hit on a coonter, an' Guid bliss me, athin twartree meenits he wis rampin an' boilin' athoot a spunk near him. (Gray 1933, also a further 3 tokens of hit=tea-kettle and 3 of he/him=tea kettle).
As noted both by Wagner (2005) and Hernandez (2011), gendered pronouns are not commonly applied to mass noun referents. On Hernandez' animacy hierarchy (2011: 85), 99.7% of inanimate mass nouns were found to occur with the neuter pronoun, none have the feminine pronoun, and only 0.3% occur with the masculine pronoun. Similarly, in Shetland dialect, Robertson and Graham (1991: 2) note the neuter gender's 'use in naming formless objects, e.g., girse, hey, mold'. Some examples from both Orkney and Shetland can be seen below:
10a: I tink gin I could get a' the bere an' maut 'at gangs oot o' Orkney tae the king, I was try a tullye for hid gin I thought I could get hid bae fechtin' for hid. (Orkney, Dennison 1880)
10b: An' whin he got de sheul o' dirt i' his face, he teuk hid for the enemy's shot. (Orkney, Dennison 1880)
10c, hot water: ... man dey wir a pooer o' haet water rinnin' oot o'yon wirm-aetin laek thing [...] bit [...] whin I stappid up ee side wi' a fael hit juist cam as fast oot da tidder side. (Shetland, Gray 1933)
10d, 'it all': ... git doo hit aa stowed in eft. (Shetland, Gray 1933)
In 10d, 'hit aa' refers to a whole list of items bought from a shop ('a iron bed, twa muckle pots, an' some lame'), which collectively make a mass of luggage.
Abstract noun referents are difficult to classify on Hernandez' animacy hierarchy, because as Wagner (2005: 348) notes, count nouns can be both concrete and abstract so that the two continua count-mass and concrete-abstract interact, resulting in a three-dimensional model. However, Wagner chooses to simplify the situation by referring to concrete count nouns as 'count nouns' and abstract count nouns as 'abstract nouns'. In any case, abstract nouns are found to be usually neuter in both spoken and written Standard English, as well as in the traditional West Country dialect (Wagner 2005: 349). In rare circumstances, abstracts can also be feminine in spoken and written Standard English. In the Newfoundland material considered by Wagner, in addition to being neuter, abstract count nouns can in rare cases also be masculine by extension of the rule that concrete count nouns can be masculine, and occasionally also feminine by extension from Standard English. In the Orkney and Shetland material, however, there are in addition to neuter abstracts also examples of both feminine and masculine, as shown below:
11 a, pardon: An' a' body wus blide for his sake whin de parden cam'--Deil con dem t'anks wha keepid hid ap t'ree weeks eftir hid cam' tae Kirkwa' (Orkney, Dennison 1880)
11b, death: I find me daeth wirkan' i' me heid. Hid's wirkan' doon; an' whin hid comes tae the he'rt, hid's a' ower wi' me! (Orkney, Dennison 1880)
11c, sermon: An' Mr Coventry herd o' the chaepal sermon; an' he praeched on the neest Sabbath, fae the text,--I ken her weel, for me fether heard the sermon, an' said the text minny a time,--'Jeudge no' that ye be no' jeudged, for wi' what jeudgment ye jeudge ye sail be jeudged.' (Orkney, Dennison 1880)
11d, job: Du's lucky ta hae sic a job, she said. Mony anidder lass wid gie her back teeth for ane laek him. (Shetland, Graham 1987)
11e, tune: He's a graand tune, whidder it was da deevil or da minister at set him on. (Shetland, Graham 1987)
11f, law or burn: Da laa is laek da burn I da hill: he taks his coorse an you canna stop him (Shetland. Graham 1987)
In 11f, it is not very clear whether him refers to the law or the burn (stream). The 'sermon' in 11c is a good example of what Wagner means when saying that abstract nouns can also be count nouns. In that respect, it is perhaps not surprising that it can take on the feminine gender like concrete count nouns generally can in Orkney dialect.
Likewise, in 12a-c from Shetland, the use of the masculine pronoun with abstract count nouns could perhaps be seen as an extension of the general rule that concrete count nouns are most often masculine in this dialect.
Having shown a range of examples of gendered pronoun usage in the Orkney and Shetland dialect material, I will now turn to the discussion, where the five hypotheses outlined in the introduction will be considered.
Starting with the null hypothesis, that there is no difference between grammatical gender in Orkney and Shetland dialects and spoken Standard English in general, this hypothesis can be discarded. As shown above, both Orkney and Shetland employ gendered pronouns to a much greater extent than Standard English, including the 'spoken standard' system documented by Wagner (2005), discussed above. A more open question is whether or how much Orkney and Shetland differ from other spoken varieties of Scots. Purves' Scots Grammar (2002: 21-34) does not mention grammatical gender in connection with nouns and pronouns. Writing specifically on northern and insular varieties of Scots, Millar (2007) only mentions gender under Shetland, as quoted above. The kind of broad ranging survey and systematic analysis which Wagner and Hernandez provide for the use of gendered pronouns in English dialects is currently lacking with respect to Scottish dialects.
Starting therefore with hypothesis 1, one needs to examine the possibility that the Orkney and Shetland dialects retain a gender system from Nom. In order to do so, it will be assumed here that genders in Nom would have matched those of Old Norse, given that modern Norwegian still largely keeps the Old Norse gender assignments intact despite other significant grammatical changes in the language. Among the tokens from the dialect text material, only those can be used in which the referent of the pronoun is a noun that has a direct Old Norse cognate. Marwick (1929: xxx) makes a mistake when he writes:
He [Walter Traill Dennison] makes geo also feminine, which might argue a faint memory of the old gender of gja, but when we find him making rock masculine (O. N. berg, n), our confidence is seriously shaken.
Here, Marwick equals 'rock' with 'berg' despite these not being cognate words. A word does not necessarily need to have the same gender as its synonym. It is easy to find examples such as the modern Norwegian tue, f. and klut, m., both meaning a cloth, but being of different grammatical genders.
Going through the referents from the sentences quoted above where there is a direct Old Norse cognate, we find (with reference to the Oxford English Dictionary, Heggstad, Hodnebo and Simensen 1975: Norron Ordbok; Dictionary of the Scots Language; Bosworth-Toller Anglo-Saxon Dictionary and MacBain: Etymological Dictionary of the Gaelic Language):
Ball, f., Old English *beall, Old Norse bollr, m., Gaelic bala, m.
Barrow, m. Old English bearwe, f. Old Norse barar, f. plural only, Gaelic bara, m.
Bear, m., Old English bera, m., Old Norse bjorn, m. (Gaelic beithir, f. for any wild beast is recorded as being related to bear by Armstrong, although there is some doubt about this).
Bed, n. Old English bed, n. Old Norse bedr, m. no Gaelic cognate.
Brig, m., Old English brycg, f., Old Norse bryggja, f., no Gaelic cognate.
Burn, m., Old English burna m., burne f., and burn f., (apparently not distinguished in sense), Old Norse brunnr, m., Gaelic burn, m. borrowing from Scots.
Cat, f. Old English cat, m., (also catte, f.: a she-cat). Old Norse kqttr, m. Gaelic cat, m.
Cubbie, f., Old English cyf f., cognate with Icelandic: kupa, f.?, no Gaelic cognate.
Death, n., Old English deap, m., Old Norse daudi, m. (also daudr), no Gaelic cognate.
Dirt, n., not known in Old English, Old Norse drit, n. (Uncertain, as the text refers to 'sheul o' dirt'), no Gaelic cognate.
Eel, m., Old English ael, eol, m., Old Norse all, m., no Gaelic cognate.
Egg, f., Old English aeg, (aeig), n., Old Norse egg, n. Gaelic ugh, m.
Fiddle, m., Old English *fi[??]ele, f., Old Norse fi[??]la, f., Gaelic fidheall, f.
Fire, m., Old English fyr, n., Icelandic furr, m., no Gaelic cognate.
Fit (foot), m., Old English fot, m., Old Norse fotr, m., no Gaelic cognate.
Geo, f., no Old English cognate, Old Norse gja, f., Gaelic geodha, m. borrowed from O.N.
Handle, n. Old English handle, f., Old Norse handar-hald, n., no Gaelic cognate.
Heid/head, m., Old English heafod, n., Old Norse haufu[??], later hqpfu[??], n., no Gaelic cognate.
Hen, f., Old English henn, f. Old Norse hona, f. no Gaelic cognate.
Hide, n., Old English hyd, f., Old Norse hu[??], f., no Gaelic cognate.
Hoose (house), m., Old English huss, n., Old Norse hus, n., no Gaelic cognate.
Horn, n. and m., Old English horn, m., Old Norse horn, n., Gaelic corn, m.
Iron, m. Old English isen, n. Old Norse jarn, n. Gaelic iarann, m.
Kettle, m. and n., Old English cetel, m. Old Norse ketill, m., no Gaelic cognate.
Kirk, m. in dialect text, f. in Roberson and Graham's grammar, Old Norse kirkja, f. Neuter in Greek, but feminine in the West Germanic languages. No Gaelic cognate.
Kist (chest), m., Old English cest, cist, cyst, f., Old Norse kista, f., Gaelic ciste, f.
Knife, f., and n. Late Old English cnif (11th cent.), m., Old Norse knifr, m., no Gaelic cognate.
Lamb, m. and n., Old English lamb, lambor (lqmb, lqmbor), Iemb, n., Old Norse lamb, n., no Gaelic cognate.
Law, m., Old English lagu, f, prehistoric Old Norse *lagu (Oxford English Dictionary), Old Icelandic lqg, f. 'In the sense of law, fem in Old Icelandic and Old Norwegian, neu in Old Swedish' (Oxford English Dictionary), Gaelic lagh, m. (Uncertain, as the text is ambiguously referring to both law and burn).
Lug, f., modern Norwegian: lugg (forelock), m. 'Of Scand. origin, in an original sense of 'something that can be pulled or laid hold of, an appendage' (Dictionary of the Scots Language). No Old English or Gaelic cognate.
Maut (malt), n., Old English mealt, n., Old Icelandic malt, n. (Uncertain, as the text refers to 'bere and maut'). No direct Gaelic cognate.
Nail, n., Old English naegel, m., Old Norse nagli, m., Gaelic iongna, f. (hoof, but ultimately cognate through an Indo-European word for claw, fingernail--see for example Latin unguis and Old Irish ingen).
Sail, m., Old English seg(e)l, n. and m., Old Norse segl, n., Gaelic seol, m. borrowed from Anglo-Saxon in Old Gaelic period 600-900 AD (Baoill 1997: 551).
Silk, f., Old English sioloc, seoloc, etc. (for earlier *siluc), m., Old Norse silki, n. Gaelic seiric, f.
Staff, n., Old English staef, m, Old Norse stafr, m., Gaelic steafag/steabhag, f. borrowed from English.
Stone, m., Old English stan, m, Old Norse steinn, m., no Gaelic cognate.
Teeble (table), f., Old English tabule, f., Old Norse tabola, f. and tafl, n., no Gaelic cognate.
(Telegraph-) pole, m. Old English pal, m. Old Norse pall, m., Gaelic poil, m.
Tune, m., no entry in Old English, Middle English ton, Old Norse toni, m., Gaelic toin, f.
Water, n. Old English waeter, n. Old Norse vatn, n. Gaelic: McBain (1911: 387) suggests an etymological connection via the ablaut-variant *ud- to Gaelic uisge, m.
Checking for correspondences between the dialect gender and the gender in Old Norse, and bearing in mind that some words are recorded with multiple genders, yields the distribution shown in Table 1. In total, 20 out of 45 words match their Old Norse gender. If the assignment had been random, one would expect a strike rate of about 0.33. The strike rate here is 0.44. Although slightly higher, it is not enough to conclude that there is a genuine retention of Old Norse genders. Removing the uncertain malt, dirt and law gives a strike rate of 0.41. An examination of a larger quantity of dialect text would be desirable.
Being essentially dialects of Scots (albeit with a Norn substrate), any remnants of a grammatical gender system in the Orkney and Shetland dialects also need to be discussed in the context of what happened to the grammatical gender system which existed in Old English. Turning to hypothesis 2, then, we may ask if the strike rate of dialect genders matched to Old English genders is any higher than those for Old Norse. Using the same words, but with their Old English genders, gives the correspondences shown in Table 2:
Table 2: Dialect and Old English gender correspondences Table 2 Masc. in 0. E. Fem. in 0. E. Neu. in 0. E. Masc. in (9) Stone, sail, (7) Fiddle, (6) Heid, fire, dialect text fit, eel, bear, kist, brig, hoose, sail, burn, horn, kirk, law, burn, lamb, iron pole, kettle barrow Fem. in (3) Silk, knife, (3) Teeble, (3) Egg dialect text cat, cubbie, hen Neu. in (6) Nail, horn, (2) Hide, handle (4) Malt, lamb, dialect text staff, death, bed, water kettle, knife
The strike rate here is 0.37, which is almost exactly the 1/3 one would expect from a random assignment. The case for Orkney and Shetland dialect having retained Old English genders is therefore no better than that of them having retained Old Norse genders.
As a further test, it would be interesting to consider those words where the gender is different in Old Norse and Old English. This applies to the following four words in the selection, with a further three which have more than one gender in Old English or Old Norse:
Bed, n. Old English bed, n. Old Norse bedr, m.
Fire, m., Old English fyr, n., Icelandic furr, m.
Horn, n. and m., Old English horn, m., Old Norse horn, n.
Silk, f., Old English sioloc, seoloc, etc. (for earlier *siluc), m., Old Norse silki, n.
Burn, m., Old English burna m., burne f., and burn f., (apparently not distinguished in sense), Old Norse brunnr, m.
Sail, m., Old English seg(e)l, n. and m., Old Norse segl, n. Teeble (table), f., Old English tabule, f., Old Norse tabola, f. and tafl, n.
In the case of silk, the Orkney dialect word is feminine, the Old English is masculine and the Old Norse neuter. The Orkney dialect word therefore follows neither Old English nor Old Norse. In the case of fire, the word is recorded as masculine in the Shetland dialect, which matches Old Icelandic but not Old English. In the case of horn, the Orkney dialect matches gender of Old Norse (neuter) while Shetland matches Old English (masculine). In the case of sail, it matches one of the two possible genders in Old English, while burn matches Old Norse and one of two possible genders in Old English. Teeble matches Old English and one of two possible genders in Old Norse (borrowed from Latin in both languages).
In consideration of hypothesis 2, one must ask whether it is possible that the variety or varieties of Scots which arrived in Orkney and Shetland still retained grammatical gender. This, however, raises the question: When did Scots arrive in Orkney and Shetland, and when did the koineisation take place, allowing the various Scots source dialects to consolidate into Orkney Scots and Shetland Scots? I do not have the scope in this article to attempt to answer this question (but see for instance Millar (2008) and Ljosland (2012) for a discussion). Marwick reckons that '(...) the pledging of the isles to Scotland in 1468 merely accelerated a process that had been going on for almost a century previously' (1929: xxii), which suggests a tentative date of about the 1370s for the beginning of the Nom-to-Scots language shift in the Northern Isles. By that time, Scots would no longer have contained any recognisable gender-specific noun morphology, as 'this system began to collapse in Northern [English] texts from the mid tenth century on, reaching fruition in South-Eastern texts around the middle of the fourteenth century' (Millar 2002: 294). The spread of the loss of grammatical gender starts in the North of England before spreading south. Grammatical gender is therefore likely to have been already lost in Scots before this language became established in Orkney or Shetland (1).
I now turn to hypothesis 3: that the gendered pronoun systems in Orkney and Shetland Scots retain features from the gender system of Gaelic. Wales (1996: 138) does not go as far as suggesting any direct transfer, but opens up to the possibility of 'influence from either Celtic or Norse languages'. Gaelic was never a community language in Orkney or Shetland, but would of course have been spoken by single individuals or families moving throughout history to the Northern Isles from Gaelic-speaking areas. This could of course cause some influence, but a more likely scenario is that Orkney and Shetland Scots resemble the kind of English or Scots spoken in Gaelic or ex-Gaelic speaking areas. The very distant familial relationship between Scots and Gaelic makes it difficult to look for cognates. However, in those cases where a Gaelic cognate can be identified, no clear pattern of correspondence emerges, as can be seen in Table 3. One must also bear in mind that Gaelic has no neuter gender.
A more interesting pattern emerges when examining the two archipelagos separately. If gendered pronouns were something which has been retained as a substrate from Norn, and Norn retained genders from Old Norse, one would expect the grammatical genders in Shetland dialect to correspond to those in Orkney dialect. In Table 4, tokens from Orkney and Shetland are compared.
As the table shows, the default gendered pronoun in Orkney seems to be the feminine when the referent is a count noun with the exception of animals, whereas masculine is the default when the referent is weather, atmospheric conditions, tides, times, and seasons. In Shetland, masculine seems to be the default gendered pronoun for count nouns, except for animate nouns, vehicles and machinery. Further evidence for this observation can be sought by considering those words which have been introduced in comparatively modern times (2) and therefore are unlikely to preserve any gender assignment from Old Norse or Old English. Such words are for example, from Shetland: croon bevel (n., end of eighteenth century (3)), Electric Faacom-cleaner (n., twentieth century), soo 'waster (m., nineteenth century), cravat (m., seventeenth century), and road roller (f., nineteenth century). And from Orkney: guinea (f., seventeenth century). As these examples show, gendered pronouns have been productive at least into the nineteenth century. Soo 'waster, m. shows that this observation applies even when one excludes vehicles and machinery. Cravat and guinea are also relatively new words with gendered pronoun references, although these are from the end of period when Norn was still spoken by some members of the community. The examples also seem to corroborate the idea that for inanimate count nouns, except vehicles and machinery, the default gendered pronoun in Shetland is masculine (soo 'waster, m.; cravat, m.) and in Orkney feminine (guinea, f.). It would be interesting to test, on a larger data set, whether pronominal references to words that have been introduced to Shetland and Orkney dialect in the twentieth century are more likely to follow the practice in spoken Standard English. At this point, however, such a data set is not available. Using Wagner's template, the Orkney and Shetland systems could be described as in Table 5:
Table 5: Gendered pronouns in Orkney and Shetland Gendered pronouns in Orkney and Shetland Humans Masculine or feminine Animals Neuter, masculine or feminine (not always according to sex) Count nouns Shetland: Masculine. Orkney: Feminine. Both areas: also neuter Mass nouns Neuter Abstract nouns Neuter, however, abstract count nouns may follow the rule for concrete count nouns
In order to evaluate hypotheses 4 and 5, we need to consider whether the systems we have now established for Orkney and Shetland are idiosyncratic developments in these areas, or whether they can be understood with reference to other dialectal varieties of English.
The Shetland system resembles the system of the traditional, rural dialects of south-western England, where the masculine pronoun can be applied to most count nouns (Wagner 2005 and 2003). However, in this light the Orkney dialect's use of the feminine pronoun in reference to count nouns seems remarkable, as this usage far exceeds any 'non-referential she' or 'personification' in Standard English. It is interesting, however, to note that Grant and Dixon (1921: 98) remark that 'highlanders are fond of the feminine pronoun for all genders', giving three examples of such practice, including: 'She was in the garden, an' she saw the rabbit, an' she took a stane an' flung 'er at 'er and kilt 'er'--here using the feminine pronoun in reference to a male servant (human, but not according to sex), a stone (inanimate count noun) and a rabbit (animate count noun). Perhaps the Orkney dialect's generalisation of the feminine pronoun to most count nouns could have something to do with the similar tendency of generalising the feminine pronoun in the Highlands?
Lastly, Orkney dialect's use of the masculine pronoun with reference to the subcategory of weather and other atmospheric conditions, tide, time and seasons seems remarkable. However, there might be an explanation for this, as I shall outline in the conclusion.
This article has shown that the dialects of Scots spoken in Orkney and Shetland employ gendered pronouns in a way that differs from written and spoken Standard English. Because of the history of language shift in the Northern Isles from Old Norse/Norn to Scots, the hypothesis was examined that Orkney and Shetland Scots retain remnants of the grammatical gender system found in Old Norse. This hypothesis had to be discarded, however, as the match between Orkney and Shetland genders and Old Norse genders amounted to little more than what one would have got from random assignment. Another argument against the idea that Norn genders should have been retained is the observation that Orkney's and Shetland's systems are mutually different. That an essentially Old Norse grammatical gender system should be retained in Orkney or Shetland Scots is perhaps not a tenable expectation for another reason, too: the encounter between Old English and Old Norse is precisely what is thought to be 'the original (if certainly not the sole) catalyst for the speed of the developments' leading to the simplification of the noun morphology in English (Millar 2000: 43). Indeed, in his brief discussion of the phenomenon, Millar concludes:
Rather less attractive is the supposed survival of grammatical gender in the use of he or she with nouns ostensibly following the Nom pattern. It is not impossible that such a cross-over should take place, although the length of time since the last native speaker of Nom spoke the language [...] makes the survival of such a limited pattern rather unlikely (Millar 2012: 20).
The present investigation has also explored the alternative hypothesis that gendered pronouns in Orkney and Shetland retain genders from Old English, as these dialects are conservative varieties of Scots. However, this hypothesis was also discarded, partly because grammatical gender, with all probability, died out in Old English/early Scots before its arrival in the Northern Isles, and partly because the present investigation failed to produce evidence of gender correspondence with Old English beyond that of random assignment.
As an alternative explanation, Millar (2012: 20) suggests that '[...] the Northern Isles system could be seen as an extension of the common habit among some speakers of English to give (normally female) sex to inanimate objects like cars or boats. ' Indeed, boats and ships are referred to as feminine in the dialect text material, along with other vehicles and machinery. However, in the Orkney dialect material the use of the feminine pronoun extends much beyond this, and its use to some extent resembles what Grant and Dixon (1921) describe as a very wide application of the feminine pronoun in Highland varieties of English. In Shetland, however, it is the masculine pronoun which is widely generalised, making any connection with either Old Norse, Old English, Highland English or personification of inanimate objects modelled on English difficult. The distribution of the masculine pronoun in Shetland dialect does, however, resemble Wagner's (2005) data from traditional rural dialects of South-West England, and might therefore not be as surprising as it first appears.
There is, however, one semantic field where there might be a case for the Orkney dialect having retained a connection with Old Norse gender, and that is regarding weather phenomena being referred to as 'he'. Barnes (1984: 364) points out that '[t]he use of he to describe meteorological phenomena, e.g. Orkney He's gaan tae blaw, I doot, also has parallels in Scandinavia, especially in the Faeroes and certain parts of Norway. ' Eriksen, Kittila and Kolehmainen (forthcoming) observe that 'quite intriguingly some languages employ a marked human pronoun as an Meteorological] Predicate] expletive. For example, in Icelandic, Faroese and some Mainland Scandinavian dialects a pronoun meaning 'he' can be used as a MP expletive.' Eriksen, Kittila and Kolehmainen continue: 'Even more intriguingly, in all these varieties of North Germanic, an unmarked neuter pronoun of the it-type (e.g. ta[??] in Faroese) is used in all other expletive contexts, like presentational sentences, showing that the marked masculine pronoun is licit only with MP's.' Considering that 'she' is so widely generalised in the Orkney dialect, the fact that weather is referred to as 'he' does strike one as significant. With Orkney being an archipelago in the North Atlantic which is exposed to much weather, one could perhaps theorise that the weather might be personified. However, 'she' being the preferred pronoun for personification in English weakens the case for such an interpretation. The presence of 'he' as the pronoun specifically referring to weather phenomena in Icelandic, Faroese and some Mainland Scandinavian dialects makes plausible a case for such a structure having been present in Norn as well. If there is indeed any traceable Norn substrate in Orkney and Shetland's gendered pronoun systems, this is it.
Anderson, B. R. 1888. Broken lights: poems and reminiscences of the late Basil Ramsay Anderson. Edinburgh: R. & R. Clark; Lerwick: C. & A. Sandison.
Angus, J. S. 1877. Eels. Collected in Echoes from Klingrahool. Poems: mostly in the Shetland dialect, with a glossary. Lerwick: T. & J. Manson, 1926.
Armstrong, R.A. 1825. A Gaelic Dictionary': in two parts I. Gaelic and English. --II. English and Gaelic. London: V. Duncan.
Baoill, C. 6. 1997: 'The Scots-Gaelic Interface'. In Jones (1997): 551-68.
Barnes, M. 1984: 'Orkney and Shetland Norn'. In Trudgill (1984): 352-366.-1998. The Norn Language of Orkney and Shetland. Lerwick: The Shetland Times.
Bosworth-Toller Anglo-Saxon Dictionary online edition bosworth.ff.cuni. cz/030777, accessed 26/3/2012.
Costie, C. M. 1976. The Collected Orkney dialect tales of C.M. Costie. Kirkwall: Kirkwall Press.
Dahl, O. 2000 'Animacy and the notion of semantic gender'. In Untersbeck and Rissanen (2000): 99-115.
Dennison, W. T. 1880. The Orcadian sketch-book: being traits of old Orkney life written partly in the Orkney dialect. Kirkwall: W. Peace & Son.
Eriksen, P. K.; Kittila, S. and Kolehmainen, L. (in press). 'The world is raining: Meteorological predicates and their subjects in a typological perspective.' Accepted, to appear in: M-L. Helasvuo & T. Huumo (eds.): Canonical and non-canonical subjects in constructions. Amsterdam: Benjamins.
Fisiak, Jacek (ed.). 2002. Studies in English Historical Linguistics and Philology: A Festschrift for Akio Oizumi. Studies in English Medieval Language and Literature 2. Bern: Peter Lang.
Graham, J. J. 1987 Shadowed Valley. Lerwick: The Shetland Publishing Company.
Grant, W. and J. M. Dixon. 1921: Manual of modern Scots. Cambridge: Cambridge University Press.
Heggstad, J., F. Hodnebo and E. Simensen. 1915.Normn ordbok. Oslo: DetNorske Samlaget.
Hernandez, N. 2011. 'Personal Pronouns'. A Comparative Grammar of British English Dialects. D. K. a. M. E. S. N. Hernandez. Berlin/Boston, De Gruyter Mouton: 53-191.
Jakobsen, J. 1928-32. An Etymological Dictionary of the Norn Language in Shetland. London: D. Nutt (A.G. Berry).
Jones, Charles (ed.) 1997. The Edinburgh History of the Scots Language. Edinburgh: Edinburgh University Press.
Knooihuizen, R. 2005. 'The Norn-to-Scots language shift: another look at sociohistorical evidence'. Northern Studies 39: 105-117.
Ljosland, R. 2012. 'The Establishment of the Scots Language in Orkney'. New Orkney Antiquarian Journal 6: 65-80.
Marwick, H. 1929. The Orkney Norn. Oxford: Oxford University Press.
McArthur, T. (ed.) 1992. The Oxford Companion to the English Language. Oxford: Oxford University Press.
McBain, A. 1911 .An Etymological Dictionary of the Gaelic Language. Stirling: Eneas MacKay. Available online at archive.Org/stream/etymologicaldict00macbuoft#page/nl0/mode/1up accessed 23/8/2013.
Millar, R. M. 2000. System collapse, system rebirth: the demonstrative pronouns of English, 900-1350 and the birth of the definite article. Oxford/New York: Peter Lang.
--2002. 'After Jones: some thoughts on the final collapse of the grammatical gender system in English.' In Fisiak (2002): 293-306.
--2007. Northern and Insular Scots. Edinburgh: Edinburgh University Press.
--2008. 'The origins and development of Shetland dialect in light of dialect contact theories'. English World-Wide 29(3): 237-267.
-- 2012. 'The death of Orkney Norn and the genesis of Orkney Scots'. Scottish Language 29: 16-36.
Oxford English Dictionary online edition www.oed.com accessed 26/3/2012. Purves, D. 2002: A Scots Grammar. Scots Grammar and Usage. Edinburgh: The Saltire Society.
Robertson, T. A. and J. J. Graham. 1991. Grammar and usage of the Shetland dialect. Lerwick, Shetland Times.
Trudgill, Peter (ed.) 1984. Language in the British Isles. Cambridge: Cambridge University Press.
Untersbeck, B. and M. Rissanen, (eds.) 2000. Gender in Grammar and Cognition. Trends in Linguistics. Berlin/New York, Mouton De Gruyter.
Wagner, S. 2003. Gender in English Pronouns--Myth and Reality. Doctoral dissertation, Albert-Ludwigs-Universitat zu Freiburg im Breisgau.
--2005. 'Gender in English Pronouns'. In Bemd Kortmann, Tanja Herrmann, Lukas Pietsch and Susanne Wagner (eds.) A Comparative Grammar of British English Dialects. Berlin ; New York Mouton de Gruyter.
Wales, K. (1996): Personal pronouns in present-day English. Cambridge [England]; New York, Cambridge University Press.
Zoega, Geir T. (2004): A Concise Dictionary of Old Icelandic. Toronto/Buffalo/ London: University of Toronto Press in association with the Medieval Academy of America.
University of the Highlands and Islands
(1) I am grateful to Dr Robert McColl Millar for his personal communication on this matter 13th March 2012.
(2) I am grateful to Dr Mercedes Durham for this suggestion.
(3) The dating information is from Oxford English Dictionary and refers to the earliest source referenced there, which might of course not coincide with when the word was first used in Shetland or Orkney.
Table 1: Dialect and Old Norse gender correspondences Table 1 Masc. in O. N. Fem. in O. N. Neu. in O. N. Masc. in (9) Stone, tire, (6) Fiddle, (6) Fleid, dialect text fit, eel, bear, kist, brig, hoose, sail, tune, burn, kirk, law, lamb, horn, iron pole, kettle barrow (f. pl.), Fem. in (4) Ball, knife, (4) Teeble, (2) Silk, egg dialect text lug, cat cubbie, geo, hen Neu. in (6) Nail, staff, (1) Hide (7) Horn, maut, dialect text death, bed, dirt, lamb, horn, kettle, knife handle, water Table 3: Dialect and Gaelic gender correspondences Table 3 Masc. in Gaelic Fem. in Gaelic Masc. in Burn, m., burn, m. Bear, m., beithir, f. dialect text Law, m., lagh, m. Fiddle, m., fidheall, f. Sail, m., seol, m. Kist, m., ciste, f. Iron, m. iarann, m. Tune, m., toin, f. Pole, m. poil, m. Fem. in Ball, f., bala, m. Silk, f., seiric, f. dialect text Egg, f-, ugh, m. Geo, f., geddha, m. Neu. in Horn, n., corn, m. Nail, n., iongna, f. dialect text Water, n., uisge, m. Staff, n., steafag, f. Table 4: Comparing distributional tendencies of gender assignment in the Orkney and Shetland dialect texts. Masc. in Orkney Gale, [other weather, atmospheric conditions, tide, time, seasons"], whale's head, polar bear, stone Fem. in Orkney Writing, skin, guinea, gun, ball, silk, knife, table, coat, peat, egg, cubbie, cap, pool, geo, sermon, lug Neu in Orkney Nail, hide, powder, horn, hood, staff, hand, 'here and maut', 'sheul o' dirt', pardon, death Masc. in Shetland Gale, [unspecified low temperature], chair, reaping hook, Beltane fire, fiddle, house, 'pig' bottle, kist, sail, bridge, kirk, foot, head, eel, whale, polar bear, lamb (might be biological gender), job, tune, law/burn, burn, horn, barrow, soo'waster, cravat, telegraph pole, an iron, weather, kettle Fem in Shetland Milking cow, hen, cat, car, engine, road roller Neu in Shetland Moorie, lamb, bed, handle, crown bevel, tow, water, kettle, 'Electric Faacom-cleaner', knife
|Printer friendly Cite/link Email Feedback|
|Date:||Jan 1, 2012|
|Previous Article:||Gaelic-English bilingual children's aspectual awareness: grammaticality judgements of English stative-progressives.|
|Next Article:||Brose, Atholl Brose, Spurtle and Thivel.|