1

The Spread of the Phrasal clitic -sa in Faroese

2

Hjalmar P. Petersen

3

Abstract 1

In Faroese the phrasal clitic -sa is used in possessive constructions like Jákup=sa2 skegg

‘Jákup’s beard’. Originally the clitic could only be attached to kinship nouns or proper nouns

of persons and other nouns with a similar function, but in recent years the -sa has spread to

other host nouns like animate, human nouns. It will be shown that the process of spreading

happens according to the animacy/referentialy split, but that language replication also plays a

role. When discussing the origin of -sa it will be shown that a form like hansa ‘his’ in the

Ballad Language shows that the old genitive plural ending -a at some point in the history of

Faroese was productive as it is in South Swedish dialects.

Keywords: Animacy Hierarchy, clitic, language replication and contact, possessive

constructions, animacy/referentiality split.

1 I would like to thank D. Galbraith for correcting my English and **.

2 In the examples clitic boundaries are marked by equals sign, as required in the Leipzig Glossing Rules.

This is done both in the object language, Faroese, and in the gloss.

4

1. Introduction

The -sa clitic is first mentioned in Lockwood (1964[1977]: 106),3 who says that -sa is mostly

avoided in the literary language, a statement that is in accordance with the following citation

from the Faroese Language Committee (Føroyska Málnevndin):

Barnsliga og óformliga hvørsfalsendingin -sa, sum í Jákupsa, mammusa, ommusa, hóskar ikki í øllum

stíli, sum t.d. tá ið onkur segði "í Bresjnevsa tíð" í staðin fyri "í Bresjnevs tíð", sum er væl bæriligari

bæði í skrift og talu.

(Orðafar, 4, June 1987)

The childish and informal genitive ending -sa as in Jákup=sa mammu=sa, ommu=sa is not suitable at

all style levels, as when someone said ‘in Bresjnev=sa time = at the time of Brezhnev’ instead of ‘in

Brezhnev’s time’, which is a form that is more dignified in both the written and spoken language.

The -sa clitic is usually ignored in prescriptive Faroese grammars as in the influential

grammar of J. Dahl (1908), and it is not mentioned at all in others, like Davidsen &

Mikkelsen (1993; 2011) and Andreasen & Dahl (1997).

3 Loockwood (1964: 106) has the forms Jákupsa(r), Marjunsa(r), Ólavsa(r), Altasa(r), Tórasa(r) and Siggusa(r) with an optional word final -r. These forms, with word final -r, are unknown to the author of the article, a native speaker of Faroese, and no one I have consulted are familiar with a form like Ólavsar. The -sar is not used in spoken Faroese, but can be found in place names like Jóannesargerði ‘J. fenced-in plot of land’. These forms are normalized from Jóannesa gerði to Jóannesar- with addition of -r, to make the construction look more like a proper genitive ending as in sonar ‘son-GEN.SG.’

5

The most extensive treatment of -sa is in Staksberg (1996), who is the first to argue, to

my knowledge, that -sa is a phrasal clitic; the presentation in Thráinsson et al. (2004: 64-65)

builds to some extent on Staksberg’s analysis and description, but they have added more

examples and provide a description of the syntax of the sa-constructions in Modern Faroese.

The syntax of sa-is also discussed in Sigurðsson (2009a/b) and in Sigurðsson (2013).

In the following I will give a short overview of the -sa clitic in Faroese, see section 2.

Before I proceed to the spread of -sa in section 4, 5, 6 and 7, I will give a short overview of

possessive constructions in Faroese, see section 3. In section 8 I will present the results from

a grammaticality judgment test carried out by students at Føroyamálsdeildin (Department of

Faroese) as part of their course in Field Work in the period between Feb. and Aug. 2015.

Section 9 explores the age and origin of the clitic in counterpoint to similar patterns in other

Scandinavian dialects, and finally there is a discussion and a conclusion in section 10 and 11.

2. The -sa clitic

Faroese has lost its genitive case and has developed different strategies for expressing

possession, see section 3 Possessive Constructions in Faroese. A short Overview; one of

these strategies is a possessive construction with the phrasal clitic -sa, which, originally,

could only be attached to a kinship noun or to ‘proper names of persons and other nouns with

a similar function’ (Thráinsson, 2004: 64-65); see also Lockwood (1977: 106) and Staksberg

(1996). The examples in (1) illustrate the use of -sa in colloquial speech. I have concentrated

on examples where the nature of the host noun is clear. For more on the syntax of -sa, I refer

the reader to Thráinsson et al. (2004: 64-65; 249-251) and Sigurðursson (2013).

6

(1) a. [Mammu]=sa bilur

mother-NOM.SG=sa-car-NOM.SG

‘Mother’s car’.

b. [Ólavur]=sa bilur

Ólavur-NOM.SG=sa car-NOM.SG

‘Ólavur’s car’.

c. [Jákup har Uppi]=sa bilur

Jákup-NOM.SG there Up=sa car-NOM.SG

‘The car of Jákup who lives up in the village’.

d. Hetta er [Beintu og Róa]=sa hús

this-NOM is Beinta-OBL.SG and Rói-OBL.SG=sa house-NOM.SG

‘This is the house of Beinta and Rói’.

e. Snar=sa skál

Snar-NOM.SG=sa bowl-NOM.SG

‘Snar’s bowl’.

(Staksberg, 1996)

The example in (1a) shows that the clitic is attached to an oblique case of a weak noun like

mamma-nom./mammu.OBL. in the singular. Weak masculine nouns end in -i in the

nominative (granni ‘neighbour’), and in -a in the oblique cases in the singular:

7

granna ‘neighbour’. Weak feminine nouns end in -a in the nominative (mamma ‘mother’),

and -u in the oblique cases in the singular, mammu ‘mother’, and weak neuter nouns end in

-a in the singular, eyga ‘eye’. In (1b) the -sa is added to the nominative ending -ur, a very

common practice, although Ólav=sa seems to be possible according to Thráinsson et al.

(2004: 250), who say that although Ólav=sa may look like the accusative of Ólavur, this is

not the case since the name with the form Ólav is in the nominative.

The sentences in (1c) and (1d) illustrate that -sa is a phrasal clitic, as it is attached to

the adverb uppi ‘up’ in (1c) [Jákup har Uppi]sa and to the second noun in (1d), [Beintu og

Róa]sa although the meaning of the utterance is that it is the ‘house of Beinta and Rói’. If -sa

was a head marker, it should be possible to attach it to the head of the phrase. The

ungrammaticality of *Jákup=sa har Uppi hús ‘Jákup’s-from-up-Yonder’ shows that this is

not the case. Similarly one could expect that *Beintu=sa og Róa=sa hús was grammatical. In

this case the -sa clitic is added to the heads Beintu and Róa, and it is not possible to do so.

The example in (e) shows further use of -sa with proper nouns (a dog’s name), but

note that it cannot be added to all proper nouns. Something like *Island=sa framtíð

‘Iceland’s future’ is not heard in the colloquial language.4

Thráinsson et al. (2004[2012]) mentioned in a footnote, that -sa was heard with the

name of a hotel in the centre of Tórshavn, the capital of the Faroe Islands, and they observe

that this suggests that this type of construction may be spreading:

(2) Hetta var í Hotell Hafnia=sa tíð

4 Proper nouns refer to single words like Snar ‘Fast’, a dog’s name’, and Jógvan ‘John’. Proper names include all proper nouns (in their primary applications) as well as noun phrases as [NP Jákup har Uppi]=sa bilur lit.: Jákup ther Up’s car.

8

this was in Hotel Hafnia=sa time-DAT.SG

‘This was during the time of Hotel Hafnia’.

Petersen & Adams (2009) mention the word form teirra=sa lit.: their=sa ‘their’ with addition

of -sa to a personal pronoun and Thráinsson et al. (2004: 251) mention systir mín=sa vegna

lit.: sister mine=sa because of ‘because of my sister’ where -sa is added to a determiner

phrase that contains a possessive pronoun.

The word order is usually as in (1), that is possessor + possessed, but there are

examples with the opposite word order, possessed + possessor in Staðarnavnasavnið

(Collections of Place Names at the Department of Faroese):

(3) a. Dráttur Tummas=sa5

pulling-place.NOM Tummas=sa

‘Tumma’s slipway’.

b. Keksið Jákup=sa

fold-the.NOM Jákup=sa

‘Jákup’s fold’.

5 The noun dráttur refers to a place, where boats are pulled on land. Other examples in Staðarnavnasavnið with the word order possessed + possessor are: Trøð Tummas Petur=sa lit.: pice of land fenced in-NOM. Tummas Petur=sa ‘Tummas Petur=sa piece of fenced in land’; grynna Niklas=sa lit. shoal-NOM Niklas=sa ‘Niklas’s shoal’; Torvan Tjørnuvíks-Jógvan=sa lit.: grassy patch on a rock surface Tjørnuvík.NPLLE-Jógvan=sa, which means: ‘Tjørnuvíks-Jógvan’s grassy patch on a rock surface’; NPLLE is an abbreviation of ‘non-paradigmatic linking element’ (Petersen & Szczepaniak, 2014); it is a non-paradigmatic linking element, as the paradigmatic linking element should be -ar as vík is a strong feminine noun; ryggin Jákup=sa lit.: ridge-NOM.SG Jákup=sa ‘Jákup’s ridge’; Trøð Tummas Petur=sa ‘Piece of land fenced in-NOM.SG Tummas Petur=sa, which means: ‘Tummas Petur’s piece of land that is fenced in’; skarðið við Steini Magnus=sa lit.: pass-the-NOM at stone-DAT Magnus=sa ‘The Pass at Magnus’s Stone’.

9

The word order possessed + possessor is in other place-names like these two from

Sandavágur: Seiðaberg Magnus Heimara lit.: fishing-place on the shore Magnus-ACC

Nearest-to-the-Home = ‘The fishing place of Magnus that is closer to the village’ and

Seiðaberg Magnus Ytra lit.: fishing-place on the shore Magnus-ACC Outer = ‘The fishing

place of Magnus that is further out from the village’. Both examples are with accusative of

possession, as also in troyggja Einar lit.: sweater.NOM Einar.ACC ‘Einar’s sweater’ and sonur

norska kongin lit.: son-NOM.SG Norwegian-ACC.SG king-ACC.SG ‘son of the Norwegian king’,

see section 3, where there is a short overview of possessive constructions in Faroese.

In Petersen (2015) further examples with the sa-clitic are presented, where the host is

something other than a proper noun or a kinship noun. An example is Tað er SEV=sa ábyrgd

lit.: it is SEV-NOM=sa responsibility-NOM = ‘it is the responsibility of SEV’, where SEV is

the company that provides the islands with electrical power.

In the rest of this article I will present the observations in Petersen (2015), the result

of a grammaticality judgment test conducted by students at Føroyamálsdeildin (The

Department of Faroese; The University of the Faroe Islands) and on results from Sigurðsson

(2009a/b; 2013) as well as research into the ballads that Jens Christian Svabo wrote down in

the period 1781-1782.

3. Possessive Constructions in Faroese. A short Overview

Here is a very short overview of possessive constructions in Faroese. For more information I

refer the reader to Thráinsson et al. (2004) and references they cite, but it should be pointed

out already here that possession in Faroese is a field should be studied thoroughly with

10

emphasis on the historical development and the distribution in the spoken and written

language of today.

Old Norse, like Icelandic, has modifying genitives like tveggja daga vist lit.: two

days-GEN.PL food-NOM ‘food for two days’ (Faarlund, 2004: 61). Such adnominal genitives

do not exist in Faroese as shown in (4), where (b) and (c) are examples with definite

genitives.

(4) a. *Her eru ríks mans hús

here are rich-GEN.SG man-GEN.SG house-NOM.SG

‘Here is the house of a rich man’.

b. *tak húsins

roof house-DEF-GEN.SG

‘The roof of the house’.

c. hali kúgvarinnar

tail cow-DEF-GEN.SG

’The tail of the cow’.

(Thráinsson et al., 2004/2012: 63)

Although the sentences in (4) are absent from spoken Faroese and sound strange, adnominal

genitives in fixed expressions with the historical genitive exist as, for example, ársins tíð lit.:

year-DEF-GEN.SG time-NOM.SG = ‘time of the year’, landsins siður lit.: country-DEF-GEN.SG

custom-NOM.SG., ‘custom of the country’ and á mansins ævi lit.: on man-DEF-GEN.SG. life-

span-DAT.SG. ‘as long as one lives’.

11

Barnes & Weyhe (1994: 207-208) mention ríkisins ovasta umboð ‘the kingdom’s

foremost representative’ (Barnes & Weyhe, 1994: 207-208), but such constructions are not

part of speakers’ active knowledge of Faroese; they are revitalized modelled on Icelandic/Old

Norse, and are supported language contact with Danish by the Danish phrase rigets øverste

repræsentant ‘the kingdom’s foremost representative’ and they are restricted to written

Faroese.

Kinship relations are expressed as in (5), which is an example of accusative of

possession6:

(5) a. Sonur norska kongin

Son-NOM.SG Norwegian-ACC.SG king-ACC.SG

‘Son of the Norwegian king’.

b. mamma Kjartan

mother-NOM.SG Kjartan-ACC.SG

‘Kjartan’s mother’.

Barnes & Weyhe (1994: 208) note that a construction like (5b) has been losing ground in the

last 40-50 years, and they also mention that earlier it was not restricted to personal

relationships, although they do not give any examples. Weyhe (2015, personal

communication) says that in the dialect of Suðuroy one could hear accusative of possession-

constructions like troyggja Einar lit.: sweater-NOM.SG Einar-ACC.SG ‘Einar’s sweater’ and 6 The term ‘accusative of possession’ is from Loockwood (1977: 103). He mentions pápi drongin lit. father-NOM.SG boy-DEF-ACC.SG ‘the boy’s father’, mamma brúðrina lit.: mother-NOM.SG. bride-DEF-ACC.SG ‘the bride’s mother’, abbi lítla Jógvan lit.: grandfather-NOM.SG little-DEF-ACC.SG Jógvan-ACC.SG ‘little John’s grandfather’ and døtur keypmannin lit.: daughters-NOM.PL grocer-DEF-ACC.SG ‘the shop-keeper’s daughters’. In all these cases, the postnominal noun is in the accusative, and they are not recent as Jakobsen (1898-1901: XLV) has pápi gentuna lit.: father.NOM.SG. girl-DEF-ACC.SG ‘the girl’s father’.

12

bátur Petur lit.: boat-NOM.SG Petur-ACC.SG ‘Peter’s boat’. Furthermore he points to the place

name in Fugloy (an island): Hav Hálvdan Úlvsson lit.: heavy stone-to-lift-NOM.SG Hálvdan

Úlvsson-ACC.SG ‘Hálvdan Úlvsson’s stone’.7 Another example with hav ‘heavy stone to lift’

is Hav Tofta-Katrin lit.: stone-NOM.SG Tofta-Katrin-ACC.SG ‘Tofta-Katrin’s stone’ in the

village Viðareiði, and I have mentioned the place-names in Sandavágur Seiðaberg Magnus

lit.: fishing-place-NOM.SG Magnus-ACC.SG’ above which also is an example of accusative of

possession.

In spoken Faroese kinship relations like the one in (5b) are often expressed with the

preposition til + accusative, mamma til Kjartan lit.: mother to Kjartan.ACC ‘Kjartan’s

mother’. In this context it is worth noting that the proper name Kjartan in (5b) is in the

accusative instead of an expected genitive, til Kjartans, as the preposition til used to govern

the genitive, (Hamre, 1961; Thráinsson et al., 2004: 176-180; Steintún, 2013). The other

prepositions that governed the genitive are: innan ’inside’, uttan ’outside’, (í)millum

’between’, the afore mentioned til ’to’, and vegna ’because of’, which is also a postposition

(Thráinsson et al. 2004: 176-180). In the case of innan + NP and uttan + NP there is

univerbation of the preposition + noun phrase so that for example [PP innan NP veggja]] lit.:

inside walls.GEN.PL has changed status to an adverb innanveggja ‘inside’. The same has

occurred in for example uttan dura lit.: outside doors.GEN.PL to uttandura ‘outside’.

The most common way of expressing possession is with the preposition hjá ‘with’ as

in (6a). Other possessive prepositional uses are presented in (6) as well. Here at ‘to’ and til

‘to’ are used to denote kinship relations, and á ‘on’ and í ‘in’ are used to denote inalienable

possession (Thráinsson et al. 2004: 62-63):

7 This particular name has been normalized to Hav Hálvdan Úlvssonar with the historical -ar genitive

ending in Staðarnavnasavnið. Unfortunately such corrections are frequently done by some of the collectors, and therefore the collection of place-names has to be handled with care.

13

(6) a. báturin hjá mær

boat-DEF-NOM.SG at me-DAT.SG

‘my boat’.

b. abbi at dreinginum

granddad-NOM.SG at boy-DEF-DAT.SG

‘The grandfather of the boy’.

c. mamma til Kjartan

mother-NOM.SG to Kjartan-ACC.SG

‘Kjartan’s mother’.

d. hárið á gentuni

hair-DEF-NOM.SG on girl-DEF-DAT.SG

‘The girl’s hair’.

e. eyguni í dreinginum

eye-the-NOM.SG in boy-DEF.DAT.SG

‘The boy’s eye’.

f. motorurin í bilinum

engine-DEF-NOM.SG in car-DEF-DAT.SG

‘the car’s engine’.

g. takið á húsinum

roof-DEF-NOM.SG on house-DEF-DAT.SG

14

‘The roof of the house’.

Of these the hjá ‘at; with’ possessive prepositional construction is becoming more and more

frequent so that it is not uncommon to hear for example hárið hjá mær lit.: hair-DEF-NOM at

me.DAT ‘my hair’ (Teldutøka tekstasavn Føroyamálsdeildarinnar = a text database at the

Department of Faroese) for hárið á mær lit.: hair-DEF-NOM on me-DAT ‘my hair’.

Possessive pronouns usually precede the noun they modify in spoken Faroese, unless

the noun is a kinship noun. This word order is illustrated in (7) from Petersen (2010: 157),

which builds the FADAC8 database in Hamburg:

(7) a. ...umvera pápa sín

avoid father-ACC.SG his-GEN.SG

b. …mín hundur

my-NOM.SG dog-NOM.SG

‘my dog’.

Faroese has four indeclinable possessive pronouns hansara ’his’, hennara ’her’ okkara

’ours’, tykkara ’yours’ as in:

(8) a. Vit eiga ein bil. Hetta er okkara bilur

we-NOM have a car-ACC-MASC.SG This is our- PL car-NOM-MASC.SG

‘We have a car. This is our car.’

8 The FADAC database contains informal interviews with young speakers, speakers of the mid generation and speakers that are 70+. They speak Faroese and Faro-Danish. For more information on FADAC see Petersen (2010) and the following link: http://www.uni-hamburg.de/sfb538/k8daenisch.html.

15

b. Vit eiga eina gentu. Vit hjálpa okkara gentu

we have a girl-ACC-FEM.SG We help our-PL girl-DAT-FEM.SG

‘We have a girl. We help our girl’.

C. Vit eiga eitt skip. Vit síggja okkara skip

we.NOM have a ship-ACC-NEUT.SG We see our-PL ship-ACC-NEUT.SG

‘We have a ship. We see our ship’.

In Old Norse the pronoun agreed with the noun it modified (Axelsdóttir, 2002). Faroese (as

Icelandic) does not show any agreement as the examples in (8) show.

To summarize: Modifying genitive forms are not part of the grammar of speakers of

Faroese. Kinship relations are expressed with the accusative; otherwise there are different

possessive prepositional constructions of which hjá ‘at; with’ is gaining ground; ascertaining

just how far this spread has gone requires further research.

In spoken Faroese, the word order is possessive pronoun + noun (mín bilur ‘my car’)

unless the head noun is a kinship noun, where the word order is noun + possessive pronoun

(mamma mín lit.: mother mine ‘my mother’). Both word orders can be found in written

Faroese (Barnes, 2002).

How exactly labour is divided between all the possessive constructions in Faroese is a

matter of further investigation, where the diachronic and synchronic dimensions must be

taken into consideration.

4. The –sa clitic added to pronouns

16

Petersen & Adams (2009: 249) provide an example with a pronoun + -sa: teirra=sa børn lit.

their=sa children ‘their children’. This form is grammatically the same as teirra børn ‘their

children’, where teirra ‘their’ is an indeclinable possessive pronoun. Adding -sa to teirra

does not contribute anything new to meaning of the construction; both teirra børn and

teirra=sa børn have the same meaning. It should be noted that the use of -sa with this

indeclinable possessive pronoun is not uncommon in the dialect of Suðuroy. Other than

teirra=sa I have included the examples in (9), where (9d) was uttered by a 16 year old boy in

December, -14 (V. Absalonsen, personal communication).

(9) a. […] teirra=sa børn

their=sa children

‘Their children’.

(Petersen & Adams, 2009: 249)

b. [...] fyri tykkum øll=sa skyld

for you-ACC.PL all-ACC.PL=sa sake-DAT

‘For the sake of you all’.

(Sigurðsson, 2009: 8)

c. [...] og man ikki “snakkar” hinum=sa høvuðssprog

and one-NOM not talks other-DAT.SG=sa main language-ACC

‘... and you don’t talk the other person’s language ...’

(Sigurðsson, 2009: 8)

17

d. So kann mann byggja, har sum annara=sa resursir eru

so can one-NOM build there where other=sa resourses-NOM.PL are

‘Then one can build there where someone else’s recourses are’.

e. […] onnur=sa viðurskifti

other.NOM=sa business.NOM

‘other people’s business’.

(Sigurðsson, 2013)

In (9b) the preposition fyri ‘for’ governs the accusative, and this is the case of the host, where

there is agreement between tykkum-ACC and øll-ACC. The nominative form is tey øll ‘they all’

and the dative is tykkum øllum ‘them all’. In (9c) hinum ‘the other’ is in the dative. The verb

snakka ‘to talk’ is borrowed from Danish and is synonymous with tosa ‘to talk’, a verb that

governs the accusative in Faroese, as in eg tosi føroyskt Lit.: I speak Faroese-ACC.

5. The -sa clitic added to Names of Institutions /Organizations

Thráinsson et al. (2004) mentioned as an example of the spread of -sa the hotel in Tórshavn:

hetta var í Hotell Hafnia=sa tíð lit.: this was in hotel Hafnia=sa time.DAT ‘This was during

the time of Hotel Hafnia’. The example shows that the -sa clitic can be attached to hosts other

than kinship nouns or proper nouns of persons and other nouns with a similar function. In

(10) I have included Hotel Hafnia and other examples that I have heard and noticed in spoken

Faroese. Note that the second noun is síðu ‘side’ in four of the examples in (10) of which

three are with the preposition frá ‘from’: [PP frá ‘from’ noun=sa + síðu] ‘from NOUN’s

18

side’. I mention the construction [PP frá noun=sa síðu] again in section 10 Discussion, as it is

very likely that we are dealing with influence from Danish here, where the replicated Faroese

construction builds on Danish [PP fra noun-s side].

(10) a. [...] í HB=sa høgru síðu

... in HB=sa right side-OBL.SG

‘...on the right half of HB’s football field’.

(Rás 2. Fótbóltsútvarpið á ólavssøku, 2012)

b. Hetta var í Hotell Hafnia=sa tíð

this was in Hotel Hafnia=sa time-DAT

‘This was during the time of Hotel Hafnia’.

(Thráinsson et al. 2004[2012]: 64, fn. 5)

c. [...] eg kenni ikki tølini, men eg havi hoyrt frá ALS=sa síðu…

I know not numbers, but I have heard from ALS=sa side-DAT

‘I do not know the numbers, but I have heard from the Unemployment Assurance…’.

d. [...] so vilja vit frá Føroya Arbeiðsgevara=sa síðu vera við til at fíggja…

...so will we from Faroese Employer Association-OBL=sa side-DAT be with to to finance-INF

‘...then we want, from the side of the Faroese Employer’s Association contribute and finance ...’.

(N. Winther, 05/04/2013, KvF9)

9 KvF is an abbreviation for Kringvarp Føroyar = the Faroese Broadcasting Network.

19

e. So blívur tað til metingar frá TAKS=sa síðu

then becomes it to estimations-ACC.PL from TAKS=sa side-DAT

‘Then it will only be estimations what regards the Taxation Office’.

(Eyðun Mørkøre, 15/02/2103. KvF)

f. Tað er SEV=sa ábyrgd at fáa sum mest burturúr

it is SEV=sa responsibility-NOM.SG to get-INF as much out of ...

‘It is the responsibility of SEV to get as much profit as possible’.

The heading says ‘Names of Institutions/Organizations’ and all the nouns that host the clitic

are in some sense institutions/organizations. In (10a) HB is an abbreviation of Havnar

Bóltfelag, the football club of the capital, (Tórs)havn. The (b) sentence also refers to an

institution of a kind, namely Hotell Hafnia in the centre of the capital. ALS in (c) is an

abbreviation of Arbeiðsloysistrygging, that is the Unemployment Assurance, and Føroya

Arbeiðsgevarafelag ‘The Faroese Employer Association’ is also an institution, and so are the

last two examples in (10), where SEV is an abbreviation of Streymoy-Eysturoy-Vágar. These

three islands formed a company for providing electricity, and SEV is now the main provider

of electric power to households on the Faroe Islands. In (e) TAKS is an abbreviation of

Tollur-Avgjald-Kolvetni-Skattur (Customs-Toll-Hydrocarbon-Tax); this is the Taxation

Office on the Faroe Islands.

From being initially only possible with kinship nouns like mammu=sa bilur ‘mother’s

car’, proper nouns of persons and nouns with a similar function as Ólavur=sa bilur ‘Ólavur’s

car’ and the pet-names like Snar=sa, the sa-clitic has spread to nouns that represent

20

institutions/organizations. I will return to the question why this has happened in section 10

Discussion.

6. The –sa clitic attached to Animate Common Nouns

So far we have seen that the clitic is attached to names of persons and pets, kinship nouns and

names of institutions (and pronouns, see more on this in 11 Discussion). This means that the

clitic is possible with nouns high in the Animacy Hierarchy, part of which Rosenbach (2008)

represents as (11) in her discussion of English genitives:

(11) Animate Inanimate

Human N > animate N > collective N > temporal N > locative N > common inanimate N

The boy’s bike > the dog’s > the company’s Monday’s mail London’s the building’s door

collar director suburb

(Rosenbach, 2008)

In Comrie (1981) proper names are highest, and mass nouns lowest:

(12) proper names > humans > animals > inanimate > abstract nouns > mass nouns.

If we assume a priori that the spread of -sa is influenced by the hierarchy, we would expect

that it would be possible to attach -sa to animate, human nouns, and this is exactly what we

find in colloquial Faroese:

(13) a. Hví gert tú tað so ikki fyri børnini=sa skyld?

why do you that then not for children-NOM/ACC.SG=sa sake-ACC.SG

21

‘Then why don’t you do it for the children’s sake?’

(Sigurðsson, 2009: 8)

b. [...] og vissi vit koma for seint, so er einki forstáilsi fyri tí frá lærarin=sa

síðu

...and if we come too late, then is no understanding from teacher-NOM=sa side-DAT

‘...and if we are late, then there is no understanding from the teacher’s side’.

(KrF 28/10/2013)

c. Har er radarin. Tað sigur abbi er sjómaðurin=sa besti vinur

there is radar. It says grandfather is seaman.DEF-NOM=sa best friend-NOM

‘There is the radar. Granddad says that it is the sailor’s best friend’.

(KvF, 26/05/2014, boy, appr. 12 yrs.)

(d) Tað eru nógv, sum siga, at tað er næmingurin=sa egna ábyrgd

there are many who say that it is student-DEF-NOM=sa own

responsibility-NOM

tað haldi eg ikki ordiliga passar. Eg haldi, at tað er næmingurin

that think I not really is true. I think that it is student-DEF-NOM.SG and

og og lærarin=sa ábyrgd

and and teacher-DEF.NOM.SG=sa responsibility-NOM.SG

22

‘There are many, who say that it is the responsibility of the student. I

don’t think that is really true. I think that it is the student and the

teacher’s responsibility.

(M. Jógvansson, age 18).

(e) Eg havi ongantíð hoyrt tað lisið frá happarans søgu -

I have never heard it read from bullier’s history -

happarin=sa síðu, men...

bullier-DEF-DAT=sa side-DAT but

‘I have never heard it from the bullier’s history, from the bullier’s side, but’

(M. Jógvansson, age 18. Example provided by V. Absalonsen)

In (13a) børnini can formally be either the nominative or accusative case in the plural with

the definite ending -ini. In (13b/c) lærarin ‘the teacher’ and is formally in the nominative,

lærari, to which the definite ending -n is added and then -sa clitic.10 In (13c) the host

sjómaðurin ‘sailor’ is in the nominative case in the singular, -ur, to which the definite ending

-in is added. In (e) the informant corrects happarans søgu ‘the bullier’s story’ to happarin=sa

síðu ‘from the side of the bullier’ which formally is a definite noun in the nominative.

7. The –sa added to an Abstract Noun

10 Note that syntactically lærarin ‘the teacher’ can be the object of the sentence, hence accusative: eg síggi lærarin lit.: I-NOM see teacher-DEF-ACC ‘I see the teacher’ although lærarin formally is in the nominative case.

23

As described above the possessive constructions are constrained not only by animacy but also

referentiality and individuation, and this is far from noteworthy. More surprising is the

following sentence that was uttered by a 60-year-old woman who works at the Teacher

Seminar in Tórshavn (Námsvísindadeildin).

(14) Mann skal ikki hava teori fyri teoriina=sa skyld

one shall not have theory-ACC.SG for theory-DEF-ACC.SG=sa sake-ACC

‘One shall not have the theory for the sake of the theory only’.

The noun teori ‘theory’ is an abstract noun, and to add -sa to an abstract noun is unexpected,

as they are low on the Animacy Hierarchy: Proper names > human > animals > inanimate >

abstracts > mass nouns (Comrie, 1981). Before adding the -sa to an abstract noun, we would

expect it to be possible with nouns that have the feature [non-human, animate], like

*hundurin=sa skál ‘the dog’s bowl’ and inanimate nouns like *húsið=sa tak ‘the roof of the

house’. I have not come across any such examples. The example in (14) may seem surprising

if Faroese is considered in isolation, but as speakers are bilingual and fluent in Danish, (14),

is very well understood as just a replication of the Danish sentence man skal ikke have teori

for teoriens skyld ‘one shall not have theory for the sake of the theory only’ where teoriens is

the genitive form in Danish. I shall return to this use in section 10 Discussion.

8. Acceptance

Sigurðsson (2009a) conducted fieldwork in 2008 together with other linguists field-work on

the Faroe Islands in connection with the NORMS-08. The method he used was magnitude

estimation, and he collected data concerning different possessive constructions in Faroese.

24

Among these possessive constructions was the -sa clitic. The results were based on

acceptability judgment from 20 consultants from six villages/towns: Fuglafjørður (North-

Eastern Faroese); Miðvágur (Western Faroese Dialect), Klaksvík (North-Faroese), Sandur

and Suðuroy (Southern Faroese) and Tórshavn (Mid-Faroese Dialect). The youngest

participant was 15 years old, and the oldest participant was 63 years old. According to the

findings in Sigurðsson (2009a), speakers willingly accept -sa when the possessor is a proper

name or a kinship noun, but not as much when the host nouns is a company/institution like

Hotell Hafnia.

In the period between February 2015 and April 2015 four students at

Føroyamálsdeildin (the Department of Faroese) at the University of the Faroe Islands took a

course in Field Work. Part of the course was to collect data on dative, and another part was to

collect data on the -sa clitic. They tested the following three sentences:

(15) a. Tað er SEV=sa ábyrgd at veita støðugan streym

it is SEV=sa responsibility-NOM.SG to provide permanent electricity-ACC.SG

‘It is the responsibility of SEV to provide permanent electricity’.

b. Tað er í lagi frá lærarin=sa síðu, at vit fáa frí nú

it is in order from teacher-DEF-NOM=sa side-DAT.SG that we get free now

‘It is ok from the teacher’s side that we can go now’.

c. Tað er í lagi frá læraran=sa síðu, at vit fáa frí nú

it is in order from teacher-DEF-ACC=sa side-DAT.SG that we get free now

‘It is ok from the teacher’s side that we can go now’.

25

The sentence in (15a) is inspired by (10f) Tað er SEV=sa ábyrgd at fáa sum mest burturúr ‘It

is the responsibility of SEV to get as much profit as possible’, a sentence from colloquial

Faroese. The one in (15b) builds on (13b)…og vissi vit koma for seint, er einki forstáilsi fyri

tí frá lærarin=sa síðu ’…and if we are late, there is not understanding from the teacher’s

side’. We used the sentences in (15) and not the actual sentence from spoken Faroese, that is,

the sentences in (10f) and (13b) because especially the sentence vissi vit koma for seint, er

einki forstáilsi fyri tí frá lærarin=sa síðu contains both vissi ‘if’ from Danish hvis and for

‘too’ and forstáilsi ‘understanding’. As Faroe Islanders are very conscious of ‘correct’

language (Nyholm-Debess, Saxov & Thomsen, 2013), and many do not like to see Danish

words on print, we had to modify the sentence when doing the judgment test. If we did not

so, we would not get judgments only on the -sa construction, but also on the words, and this

would mean that many would judge the sentences as bad Faroese just because of the

vocabulary.

The questionnaires were distributed between four age groups: 18-20, 20-40; 40-60

and 60+, and there were men and women among the informants. In the age-group 18-20 we

had 56 informants, and there were 50 informants in the other age-groups. They had four

possibilities to choose between with regard to an answer: (1) the sentence is grammatical; (2)

the sentence is grammatical, but I would not use it myself; (3) the sentence is doubtful and (4)

the sentence is ungrammatical.

26

Figure 1. Result of the grammaticality judgment test that was done as part of the Field Work course at Føroymálsdeildin

(Department of Faroese) in the period February to April 2015. In the figure the answer to (1) and (2) are taken together

to represent ‘yes’, and (2) and (3) are taken together to represent ‘no’. Answer (1) was ‘the sentence is grammatical’; (2)

the sentence is grammatical, but I would not use it myself; (3) the sentence is doubtful and (4) the sentence is

ungrammatical. Blue is ‘yes’; read is ‘no’ and green is ‘not answered’.

The results show that the oldest generation (60+) accept SEV=sa, and lærarin=sa/læran=sa.

The others are more reluctant to accept the clitic with names of institutions and nouns with

the feature [human], but the sentences are not totally ruled out. I shall return to the results of

the judgment test in section 11 Discussion.

9. Age and Origin

One of the difficulties in Faroese language history research is the paucity of written sources,

quite the opposite of the situation in Icelandic, with an unbroken written standard from 1200

onwards. One place to look for changes and older language forms in Faroese is the Ballad

Language. It is, as the name implies, the language in the Faroese ballads, of which the oldest

date back to the Middle Ages. The ballads are collected in Corpus Carminum Færoensium

(CCF) and were published between 1941 to 2003. CCF is a normalized version of the ballads,

but philological works have been published that show the ballads in their original form as

27

they were written down from the latter half of 1700 onwards before the advent of a written

standard in 1846.

A special case is represented by the tættir ‘satirical ballads’ which is plural of the

noun táttur. They were composed for special occasions, often with ridicule in mind, and one

of these tættir is Jákupsa skegg ‘Jacob’s beard’; here we see for the first time, the -sa clitic

added to a personal name. According to Eivind Weyhe (p.c.), a specialist in Faroese Ballads,

Jákupsa skegg dates back to the mid 1700s. He bases his dating on the persons that are

mentioned in the satirical ballad.

The -sa clitic is also elsewhere in the ballad language as in the title of the ballad about

king Herod, which is called Rudsa Vysan in the manuscript (Weyhe, 2003). With normalized

spelling the form is: Rude=sa Vysan (normalized spelling: Ridisa vísan) ‘The short ballad

about Herod’. This ballad was written down by Óli Jespersen (born 1810) from Suðuroy

around 1840 (Weyhe, 2003: 37).

The examples show that the -sa clitic can be dated back to at least the middle of the

eighteenth-century, and it is linked to the decline of the genitive, which was well underway

already before 1600 (Weyhe, 2013), but we shall see, in fact the plural genitive in –a was

productive at some point in the history of Faroese, just as in South Swedish dialects

(Josefesson, 2009).

Dahl (2010) mentions -sa in Swedish dialects from Träslövsläge (Halland); among the

examples are Mormor=sa katt ‘Granny’s cat’ and Alfred=sa kärring ‘Alfred’s wife’.

According to Dahl (ibid.) the -sa clitic in Swedish dialects is a combination of genitive

-s (singular) and genitive -a (plural). A more extensive study of the a-genitive of South

Swedish dialects was carried out by Josefsson (2009); in this study she concludes that it is

necessary to differentiate between two dialects, dialect A and dialect B. Speakers of dialect A

28

use -a (the Old Swedish/Old Norse genitive plural morpheme) as a morphological marker on

names in simplex words like Erik-a ‘Erik’s’ and they use -s as a phrasal clitic: Erik i Nygård-

s ‘Eric in Nygård’s’. Speakers of dialect B use the genitive -a with different variants as a

phrasal clitic (Josefsson, 2009: 215). Among the variants she lists -a, -as and -sa with

examples like Erik från Hånger-a ‘Erik from Hånger’s’ and den gamle smen=sa ‘the old

smith’s’ (Josefesson, 2009: 216). 11

The origin of Faroese -sa is, as in Swedish, genitive -s and genitive -a, and it was the

-a that was added first to the masculine personal pronoun hans > hansa ‘his’ which extends

to feminine henn-a ‘her’, see also Josefsson (2009).

There is direct evidence for this in the ballads that J. Chr. Svabo (1746-1824) wrote

down in the period between 1781 and 1782 (Matras, 1939). In the ballad Grimar kongur í

Garðarríki ‘Grimar, the king of Novogorod’ I found nine different examples with -a added to

the personal pronoun. In verse 52, 72, 73, 86, 88, 92, 103 there are examples with hans-a lit.:

his-GEN=a ‘his’. In all the examples the following word starts with a consonant, but in the

ballad Skrímsla (verse 24) the following word starts with a vowel: Hansa er brúður… ‘his is

bride…’. This means that -a is not inserted as an epenthetic vowel in front of a consonant in

the examples from Grimar kongur í Garðarríki., c.f. also Torsteins kvæði verse 141 snarliga

uppá bakið hansa ‘abruptly up on his back’ with hansa as the last word in the sentence. This

rule out that -a is an epenthetic vowel. Furthermore the historical genitive in -a occurs in

11 I can mention that -sa is in the dialect of Bornholm (Danish) (Text 673, Øster Marie, speaker: J. P. Kuhre). In the transcript, line 23 /haɲ læ:zər på pær=sa rydʑ/ = han læser på Pær=sa ryg he-NOM reads on Per=sa back-OBL = Standard Danish han læser på Pærs ryg ‘he reads on Per’s back’. According to Wimmer (1908: 97) -sa in the dialect of Bornholm is used especially in names of males like Pærs and Pær=sa, and when the names end in -a as in Ola and genitive Olǝ=sa. Names with word final -s show up with -sa as: Hans, Jens, Anars, Mas and gen. Hansa, Jensa, Anarsa, Masa where –a is added to the genitive –s. Compare Faroese names that end in –s as Jóannesa klettur ‘Johanne’s cliff’ (Staðarnavnasavnið ‘The Collection of Place Names’). Corresponding to Faroese hansa ‘his’ and henna ‘her’ the Bornholm-dialect has hansa ‘his’ and hæna ‘her’ (Wimmer 1908).

29

henna for hennar ‘her’ in Viljorms kvæði verse 46, henna ryggur ‘her back’, and in

Fuglakvæði verse 1 in Kraaku-Korn og henna Fugl ‘K.K and her=a bird. There are other

examples of hansa as well as henna but these illustrate the forms that we find in the ballad.

Modern Faroese has hansara ‘his’ < Old Norse hans and hennara < Old Norse

hennar ‘her’, and these forms are found in Svabo’s ballads as in hansara ferð (Torbin Bekil,

verse 129) and hennara heysi ‘her head’ in the same ballad (verse 166), and I have even

found hennas in Grimur kongur í Garðaríki in í hennas kvørja Hirningun ‘in all her horns’

verse 196.

To summarize the origin of -sa: The Swedish dialect data shows that -a was used with

simplex words and -s as a phrasal marker, and that it developed into what Josefsson (2009)

labels as dialect B, where there is a mix between the two markers, which came to function as

a phrasal clitic. In Faroese the historical genitive ending –a was originally added to the

personal pronoun hans > hans-a ‘his’ and nouns ending in -s as Jóannes-a and even Herodes

> Rudis-a. The form hansa made it possible for a feminine form henna to develop (Old Norse

hennar ‘her’), and as we had Rudis-a ‘Herod’s’ and hansa ‘his’, -a could be added to Jákup-s

‘Jákup’s’ so the result was Jákups-a. The -a was furthermore added to the old genitive

feminine form hennar > hennara, which is the modern form, and as speaker at some point

had hansa, henna, hennara, a masculine form hansara could easily be created: hansa :

hennara : X => hansara ‘his’. A productive -a explains also Modern Faroese okkar-a ‘our’

(Old Norse okkar) and tykkar-a ‘your’ (Old Norse ykkar). Note the form hennas ‘her’ in

Grimur kongur í Garðaríki in verse 196. Is shows that we are not dealing with deletion of –r

in Old Norse hennar > henna, but addition of –a to henn- based on the masculine form hans-

a. Once henna has developed, the –s can be added to this form hennas, although it did not

survive.

30

Another possibility is that -sa has its origin in the possessive pronoun in the third

person sín ‘his, her’. This pronoun could presumably have an unstressed form [se] in e.g.

Jógvan sín bátur ‘John his boat’ > Jógvan-se bátur lit.: John his boat ‘John’s boat’, compare

the Garpe Genitive in Norwegian Jon sin båt lit. John his boat = ‘John’s boat’ If -sa has its

origin in sín it fits better to the fact that with repetition independent words are subject to

phonetic reduction or erosion and grammaticalization (Heine & Kuteva 2005: 80; Bybee

2007: 965). This explanation is, however, not plausible as the construction [NP sín NP] is

nowhere to be found in Faroese dialects or any written sources. Another thing that makes the

grammaticalization of an unstressed se form suspect is the problem of how to get from -e

vowel in a putative se to the actual -sa clitic.

10. Discussion of spread in Modern Faroese

The -sa clitic must originally have been used to express prototypical possessive relations such

as kin relations as Jógvan=sa mamma ‘John’s mother’, mammu=sa barn ‘mother’s child’,

body parts Jákup=sa skegg ‘Jákup’s beard’ (in the ballad from 1700), and relations

expressing legal ownership as pápa=sa bátur ‘father’s boat’, Jógvan=sa bátur ‘John’s boat’.

The Animacy Hierarchy is one of the most elementary principles in all languages, and

according to this hierarchy, nouns that denote individuals are highest and mass noun lowest:

31

(16)

Proper Names Humans Animals Inanimate Abstracts Mass Tangible Objects nouns

Animates Inanimate Proper n. Common n.

Count nouns Mass nouns

(Comrie, 1981)

The question is, whether the spread of -sa follows the hierarchy in (16). After all it is hardly

surprising that -sa is attached to nouns high in the hierarchy, as prototypical possession

typically requires human, animate nouns, c.f. the kinship relations Jógvan=sa mamma

‘John’s mother’, body part expressions like Jákup=sa skegg ‘Jákup’s beard’, and legal

ownership Hjalmar=sa pakkhús ‘Hjalmar’s storehouse’. Another question that must be

addressed is if the Animacy Hierarchy in (16) alone can explain the spread of -sa, or if the

spread is better understood as a case of possession split, that is ANIMACY/REFERENTIALITY

SPLIT (Koptjevskaja-Tamm, 2002; Rosenbach, 2008: 164). This means that both animacy and

definiteness play a role for a sentence with -sa to be felicitous.

Note that the Animacy Hierarchy as such is not the topic of discussion in this article. I

intend to use Comrie’s 1981 version, and will mention the version presented in Rosenbach

(2008) in her discussion of genitives in English in addition to Zaenen et al. (2004).

Even though the Animacy Hierarchy is a fundamental principle in all languages, the

distinctions are not and need not be the same in different languages, and the distinction can

change over time in a given language but it is presumably the case that the top and the bottom

of the hierarchy are robust.

Rosenbach (2008) does not include proper names in her discussion of animacy and s-

and of- genitives in English, but she mentions that the choice between the two genitives (the

32

s-genitive and the of-genitive) is sensitive to graduated animacy values along the scale in

(11), repeated here for the sake of convenience:

(17) Animate Inanimate

Human N > animate N > collective N > temporal N > locative N > common inanimate N

The boy’s bike > the dog’s > the company’s Monday’s mail London’s the building’s door

collar director suburb

Zaenen et al. (2004) have a subdivision of the categories animate and inanimate distinction

where animate contains: animate: organizations, animals, intelligent machines and vehicles,

and inanimate contain: concrete inanimate, non-concrete inanimate, place and time.

This can be used to explain the spread of -sa, as it spreads to brand- and company

names (organizations) in SEV=as ábyrgd ‘Sev’s responsibility’. The spread to organizations

is understandable as these are often presented as groups of humans engaging in actions that

are typically human, like pronouncement and decision making (Zaenen et al. 2004). Comrie

did include proper names in his discussion of the Animacy Hierarchy. Zaenen et al. (2004)

and Rosenbach (2008) do not have proper names, although they must be included, if we shall

explain the spread of -sa to í Titanic=sa tíð ‘during the time of Titanic’ (Sigurðsson, 2009a)

and the dog name Snar=sa skál ‘Snar’s bowl’. But not all proper names can take the -sa

clitic; *Ísland=sa framtíð ‘the future of Iceland’ is impossible.

In section 6 examples like børnini=sa ‘the children’s’ were presented, that is, the -sa

had spread to animate, human nouns. Note however that there are no examples with an

indefinite animate, human noun + -sa. A form like *børn=sa ‘children’s seems to be

impossible. This means that the Animacy Hierarchy alone cannot account for the spread of

-sa. After all, børn ‘children’ and børnini ‘the children’ have the same feature, [human,

animate]. Additionally there are examples where -sa is added to pronouns like hinum=sa ‘the

other’s’, and pronouns are not included in the Animacy Hierarchy, but pronouns are high in

33

referentiality. The spread of -sa is thus better understood if animacy and referentiality are

taken together (the ANIMACY/REFRENTIALITY SPLIT, Koptjevskaja-Tamm (2002):

(18) Animacy scale: human > animate > inanimate Referentiality/definiteness scale: pronouns > proper nouns > definite > indefinite > non-specific

(Rosenbach, 2008: 164)

This means that if the possessor is high in animacy and definiteness, addition of -sa is

possible:

(19)

Possessor Prenominal Postnominal

High in animacy/ Jákup=sa skegg ‘Jákup’s beard’ Dráttur Tummas=sa ‘Pulling place of Tummas’ definiteness Mammu=sa bilur ‘mother’s car’ Snar=sa skál ‘Snar’s bowl’ (a dog’s name) hansa dýra drotning ‘his beautiful queen’.

hinum=sa høvuðssprog ’the other’s main language’ SEV=sa ábyrgd ’SEV’s responsibility’ Sjómaðurin=sa besti vinur ’sailor’s best friend’ *hundurin=sa skál ‘the dog’s bowl’ Low in animacy/ *eitt barn=sa skyld ‘a child’s fault’ Definiteness *bilurin=sa motorur ‘car’s engine’

34

Having said this, it must be noted that Animacy and definiteness are not the only

factors at play, as it is very likely that language contact and bilingual language processing is

responsible for and propelling the spread of -sa with the abstract noun teoriina=sa ‘theory’s’.

All speakers of Faroese, who are born in the Faroe Islands, are bilingual with Faroese as their

first language, L1, and Danish as their second language, L2 (Petersen, 2010). Contact is

possibly not the reason behind the spread of -sa, but it speeds up the process.

I have not been able to find examples where -sa is added to a noun that denotes an

animal, for example *hundurin=sa skál ‘the dog’s bowl’, and nor are there any examples

with an inanimate, concrete noun as *húsið=sa tak lit.: house=sa roof ‘the roof of the house’,

but strangely enough I have heard one example where -sa was added to an abstract noun:

teoriina fyri teoriina=sa skyld ‘theory for theory=sa sake’.

The example mann skal ikki hava teori fyri teoriina=sa skyld ‘one shall not have the

theory for the sake of the theory only’ was uttered by a woman aged 60. This is the only

example with -sa added to an abstract noun, and it is strange and suspect, as there are no

examples with the clitic added to inanimate, concrete nouns and animate, non-human nouns,

that are higher in the Animacy Hierarchy. Rosenbach (2008: 155) mentions that the s-genitive

was possible in early Modern English with an abstract noun like love in: in vain I struggle

with Love’s mighty Yoke (ARCHER, 1686fane.d1), Rosenbach, 2008: 155). This is

reminiscent of teoriina=sa in Faroese, but as the stage in the process represented by

inanimate, concrete nouns is not found anywhere in the material, nor *hundurin=sa skál ‘the

dog’s bowl’, I think that there is another straightforward explanation of teoriina=sa skyld,

and i.e. the intense language contact and bilingual language processing where Faroese and

Danish are activated simultaneously. There are thus no monolingual speakers of Faroese;

they are asymmetrical bilinguals with a high level of proficiency in Danish (Petersen, 2010).

Obviously this contact situation and multilingual language processing has had its impact on

35

Faroese, and it is against this background that at least teoriina=sa skyld ‘theory’s sake’ ought

to be understood.

As in all bilingual speech L2 is always ‘on’ in the background and it has been claimed

that L2 is never completely switched off, ‘even in completely monolingual situations’

(Thomason, 2001: 53). This fact obviously affects the grammar of both L1 and L2.

According to Kroll & Dijkstra (2002) the bilingual speaker has non-selective access to words

in his/her L1 and L2. This means that if a speaker wants to use a word in his L1, say Faroese

teori ‘theory’, then there is activation of the same words in his/her L2, so that the Danish

words teori ‘theory’ is activated together with the Faroese words. Activation of

homophonous dimorphs12 eases the transfer of the Danish material to Faroese, and it is of

importance to note that the transfer need not be a direct copying of some patterns in the

model language. This means that the Model language, Danish, has a possessive construction:

mandens bil lit.: man-DEF-GEN car-NOM ‘the man’s car’, husets tag lit.: house-DEF-GEN roof-

NOM ‘the roof of the house’, or teoriens skyld lit.. theory-DEF-GEN sake-NOM ‘theory’s sake’.

Speakers of Faroese try to find an equivalent possessive construction in Faroese, and they

choose the -sa to develop the possessive construction: modifying noun + noun. The use of the

new form may be restricted and not just a copy of the model, and the calquing may take

centuries, as witnessed by the fact that we still do not find examples in Faroese with -sa

attached to common nouns in general as *húsið=sa tak lit.: house-DEF-NOM=sa roof-NOM ‘the

roof of the house’.

12 Homophonous diamorphs are words in L1 and L2 that sounds almost the same.

36

The sentence in (20) shows the lexical correspondence between Faroese and Danish,

and we see that there is a 1:1 correspondence between the languages with regard to

homophonous diamorphs, not only the -sa possessive but the whole sentence:

(20) a. Man skal ikke have teorien for teoriens skyld (Danish)

b. Mann skal ikki hava teoriina fyri teoriina=sa skyld (Faroese)

one shall not have theory-ACC for theory.ACC=sa sake

‘One shall not have the theory for the sake of the theory only’

The phrase lærarin=sa síðu is also in all likelihood an instance of replication in Faroese of

the Danish model lærerens side; I say this, because the second noun is síðu (in some oblique

case) ‘side’, and this is not the only example with síðu as the second noun. Other examples in

the material are ALS=sa síðu ‘ALS’s side’; Føroya Arbeiðsgevara=sa síðu lit.: Faroese

Employer’s side’; TAKS=sa síðu ‘Taxation Office’s side’ and lærarin=sa síðu the teacher’s

side’. The construction [noun + s + SIDE] is very common in Danish, e.g. Folketingets side

‘Parliament’s side’, and this further supports the claim that the spread of -sa is accelerated by

language contact, where DA [noun + s + SIDE] is replicated by Faroese speakers as [noun +

sa + SÍÐA].

In the judgment test, see section 9. Acceptance, we saw that people in general are

uncomfortable with the -sa clitic, and the same was evident in the statement from the

Language Committee, which directly stated that -sa was childish. This in turn makes it very

difficult to gain data on the -sa clitic in written records and to get judgments in a

grammaticality judgment test. The examples that I have mentioned are mostly examples that

one has to catch on the spot, so to speak. As -sa is so stigmatized, it is quite certain that this

has influenced the results of the judgment test, as one must keep in mind that Faroe Islanders

37

are very aware of language and discuss ‘correct’ language use all the time (Nyholm-Debess,

Saxov & Thomsen, 2013).

11. Conslusion

The spread of -sa is partly an internal change, where -sa is attached to nouns that are high in

animacy and definiteness/referentiality. Language contact and bilingualism accelerates the

spread, where a fully fledged category that is used obligatorily in the source language has a

different copy in the replica language; this copy or calque is not used obligatorily and is

rejected by prescriptive grammarians as witnessed by the quotation at the beginning of this

article.

References

Andreasen, Paulivar & Dahl Árni. 1997. Føroysk mállæra. Føroya Skúlabókagrunnur,

Tórshavn.

Axelsdóttir, Katrín. 2002. “Hvarf eignarfornafnanna okkarr, ykkarr og yð(v)arr”. Íslenskt mál

og almenn málfræði 24, 31-107.

Barnes, Michael. P. 2002. “Head + modifier or modifier + head?” In Anfinnur Johansen,

Zakaris S. Hansen, Jógvan í Lon Jacobsen & Malan Marnersdóttir (eds.). Eivindarmál.

Heiðursrit til Eivind Weyhe á seksti ára føðingardegi hansara 25 apríl 2002. Føroya

Fróðskaparfelag, Tórshavn, pp. 59-66.

Barnes, Michael P. & Weihe, Eivind.1994. “Faroese”. In Ekkehard König & Johan van der

Auwera (eds.) The Germanic Languages. Routledge, London, pp. 190-218.

38

Bybee, Joan. 2007. “Diachronic Linguistics”. In The Oxford Handbook of Cognitive

Linguistics, Dirk Geerarts & Hubert Cuyckens (eds.). Oxford University Press, Oxford, 945-

988.

Comrie, Bernard. 1981. Language Universals and Linguistic Typology: Syntax and

Morphology. Blackwell/Blackwell and Chicago University of Chicago Press, Oxford.

Dahl, Jákup. 1908. Føroysk mállæra til skúlabrúks. Tórshavn.

Dahl, Östen. 2010. Grammaticalization in the North: Noun Phrase Morpho-syntax in

Scandinavian Vernaculars. http://www.diva-

portal.org/smash/get/diva2:330145/FULLTEXT01.

Davidsen, Kári & Mikkelsen, Jonhard. 1993. Ein ferð inn í føroyskt. Føroya

Skúlabókagrunnur, Tórshavn. [Reprinted in 2011].

Faarlund, Jan Terje.2004. The Syntax of Old Norse. Oxford University Press, Oxford.

Hamre, Håkon. 1961. “The Use of Genitive in Modern Faroese”. Scandinavian Studies 33,

231-246.

Heine, Bernd & Kuteva, Tania. 2005. Language Contact and Grammatical Change.

Cambridge University Press, Cambridge.

Jakobsen, Jakob. 1898–1901. Færøske folkesagn og æventyr. S.L. Møllers bogtrykkeri,

København.

Kroll, Judith & Ton Dijkstra. 2002. “The Bilingual Lexicon”. In Handbook of Applied

Linguistics, Robert R. Kaplan (ed.). Oxford University Press, Oxford, pp. 301-321.

Orðafar. 1987. 1. árg. 4. June. Føroyska málnevndin. Tórshavn.

39

Koptjevskaja-Tamm. Maria, 2002. “Adnominal Possession in the European Languages: Form

and Function”. Sprachtypologie und Universalienforschung 55, 141-171.

Loockwood. William B. 1977. An Introduction to Modern Faroese. Føroya

Skúlabókagrunnur, Tórshavn.

Matras, Christian. 1939. Svabos Færøske Visehaandskrifter. Bianco Lunos Bogtrykkeri A/S,

København.

Nyholm-Debess, Iben; Sandra Saxov & Sóleyð Thomsen. 2013. Skal jeg tale dansk, som

bedstefar gør det, eller ordentligt?” En undersøgeles af udtalen af og holdninger til dansk på

Færøerne. BA project. University of Copenhagen, Copenhagen.

Petersen, Hjalmar P. 2010. The Dynamics of Faroese-Danish Language Contact.

Universitätsverlag Winter, Heidelberg.

Petersen, Hjalmar P. 2015. Føroysk mállæra. Bind II. Setningafrøði. (Manuscript).

Petersen Hjalmar P. & Szczepaniak, Renata. 2014. “Frá hvørsfalli til klípi”. Fróðskaparrit

61, 163-179.

Petersen, Hjalmar P. & Adams, Jonathan. 2009. Faroese. A Language Course for Beginners.

Stiðin, Tórshavn.

Rosenbach, Anette. 2008. “Animacy and Grammatical Variation - Findings from English

Genitive Variation”. Lingua 118, 151-171.

Sigurðsson, Einar. 2009a. “Genitive and possessive constructions in Faroese”. Handout. The

5th International Conference on Language Variation in Europe 25-27 of June 2009. The

University of Copenhagen, Copenhagen.

40

Sigurðsson, Einar F. 2009b. “Eignarfall og eignarliðir í færeysku”. Handout. Hugvísindaþing,

13-14 mars 2009. Háskóli Íslands, Reykjavík.

Sigurðsson, Einar F. 2013. “The syntactic phrasal ‘sa marker in Faroese” (Manuscript).

Staksberg, Marius. 1996. “Sa-possessiv”. Málting 18, 28-34.

Steintún, Bjarni. 2013. Hvørsfall. Steinrenningar og skyldskapur. BA Project.

Føroyamálsdeildin. Fróðskaparsetur Føroyar.

Svabo, Jens Christian. 1781-1782. See Matras Christian 1939.

Thomason, Sarah G. 2001. Language Contact. Georgetown University Press, Washington,

DC.

Thráinsson, Höskuldur, Petersen Hjalmar P, Jacobsen, Jógvan í Lon & Hansen & Zakaris S.

2004. Farose. An Overview and Reference Grammar. Føroya Fróðskaparfelag, Tórshavn.

Weyhe, Eivind. 2003. Í miðjum grasgarði. Rannsóknir í kvæðauppskriftum úr Suðuroy.

Føroya Fróðskaparfelag, Tórshavn.

Weyhe, Eivind. 2013. “Skomedal farin”. http://setur.fo/tidindi/trygve-skomedal-farin/.

Wimmer, Ludvig. 1908. Bornholmsk ordbog (ed. By J.C.S Espersen). Rosenkilde og Bagger,

København.

Zaenen, Annie et al. (2004). “Animacy Encoding in English: why and how”. Handout.