|
2
NORMS Føroyar
Onsdag 10. september
Individual reports from NORMS field work in the Faroe Islands: Part III
Below follow additional individual reports from Kristine Bentzen, Kristján Árnason, Janne Bondi Johannessen, Phillipp Conzett, Saija Tamminen-Parre, Höskuldur Thráinsson, and Øystein A. Vangsnes.
From Kristine Bentzen – Verb movement in embedded clauses in Faroese:
 |
| Kristine Bentzen: "The verb went that way." (Photo: Caroline Heycock) |
During the fieldwork on the Faroe Islands I investigated word order in embedded clauses (together with Gunnar Hrafn Hrafnbjargarson). More specifically, I looked at the position of the verb with respect to adverbs and negation. In general, the informants preferred not moving the verb across adverbs and negation. In my results, this preference is particularly strong in contexts with finite main verbs and negation ikki or the adverb ongantíð ‘never.’ Most informants were very reluctant to the b-example in (1), although moving a finite main verb across for example so ofta ‘so often’ was quite acceptable, as in (2):
(1) a. Hon fortaldi mær hví Ása ikki át blóðpylsu.
b. *Hon fortaldi mær hví Ása át ikki blóðpylsu.
(2) a. Jógvan skilir ikki hví Kjartan so ofta vaskar hansara bil.
b. Jógvan skilir ikki hví Kjartan vaskar so ofta hansara bil.
Based on a preliminary look at the overall data from all the locations (Tórshavn, Fuglafjørður, Klaksvík, Tvøroyri, Sandur, and Miðvágur), it looks like movement of an auxiliary (perfective or modal) in general is slightly better than movement of a finite main verb, see (3)-(4).
(3) Tey undraðust á hvi Anna hevði ongantíð drukkið kaffi.
(4) Veitst tú hví hann má altíð læna pengar frá vinunum?
Just for fun I also threw in a couple of binding tests in some of the examples, and it turns out that several informants optionally allow local binding of hansara in (2). So the examples in (2) can either mean that Jógvan does not understand why Kjartan is cleaning Jógvan’s (or some other person’s) car so often, OR that Jógvan does not understand why Kjartan is cleaning Kjartan’s car so often. (Note that these speakers also use the reflexive sína, which unambiguously yields a local binding interpretation).
The results from my fieldwork are now available in the Document Chest.
– Kristine
From Kristján Árnason:
 |
| Kristján Arnason, University of Iceland. (Photo: Johannes Myrstad Vangsnes) |
First of all I‘d like to thank the organisers for inviting me, and for the wonderful opportunity it created for me to visit the Faeroes and to gather data. This was a special happening, one of those things that can perhaps only take place under the banners of “nordisk samarbejde” (including the generous food and drink). It was great to be a part of it, and I’m sure that I’d never have managed to get the money on my own to do this!
What I did in the fieldwork sessions was to record interviews with the “natives” and discuss their views on their language and language varieties. I also had the informants read a text, which tells us about the hopeless battle of the inhabitants of Hestur against the invasion of mice. The text is to be found in Lockwood’s Introduction to Modern Faroese on page 177. I can recommend it!
The aim of these recordings was to get a systematic overview, by way of a sort of “pilot study”, of Modern Faroese phonology. I made about forty recordings, 8 to 10 minutes in length each, where the informants read the text and we discussed things like the special conditions in each of the locations and linguistic variation. Unfortunately I missed the last fieldwork session in Miðvágur, but Einar Freyr was kind enough to have four of his informants read the exiting story about the infiltration of mice into the, up to 1908, “pure” community of Hestur. And Einar even managed to get one informant to demonstrate the famous drynjing. (See sound file: drynjing2)
On the basis of these recordings it should be possible to look for and define a set of Labovian variables for a more thorough and systematic study of phonological variation in the Faroese. A way of going about this is to design a test-procedure, somewhat like the one that me and Höskuldur and others used in our survey of Icelandic phonological variation in the eighties. The variables could on one hand be “traditional” features, such as the ai vs. ɔi in veit ‘I know’, and the diphthongal vs. monophthongal pronunciation of á in bátur, variant distribution of preaspiration etc. But there are other more recent things that may be investigated, such as the reduction in the final syllables of undir ‘under’ or undur ‘wonder’, and truncation in expressions like ɛˈvɔiʧʰ for eg veit ikki etc., which seem at first glance to be more typical of the speech of younger informants. The survey could then be conducted in a way similar to what we did in Iceland with interviews aimed at eliciting examples of the phonological properties investigated.
What I found perhaps most interesting was how conscious the speakers were about the dialect differences and how willing they were to discuss them. A typical comment (particularly in Suðuroy) was that each village (bygd) had its own variety (bygdarmál). People seemed to be fairly relaxed about this, and each was fond of his own variety. Interestingly, the ideas the speakers have about their language are not always in harmony with what they actually say. (See sound file: bilr, which is from Tórshavn, where the informant and the interviewer seem to be equally confused. Please note that this was on the first day of fieldwork, and I think my Faroese improved a bit during the following sessions, which was one of the great advantages of taking part.)
Attachments:
– Kristján Árnason
From Janne Bondi Johannessen:
 |
| Janne B. Johannessen and Helge Sandøy at Matstovan hjá Elisabeth at Viðareiði. (Photo: Øystein A. Vangsnes) |
Höski and Zakaris are very often right, but not always...
I tested six informants altogether for the dialectal features that were summarised in Höskuldur and Zakaris’s very nice overview of Faroese syntax in Tórshavn 4. August 2008. Their presentation was based on earlier relevant literature on these topics (for references, see H&Z in the ScanDiaSyn Document Chest, which I put there).
I tested 40 test sentences, illustrating a number of constructions. In short, I found that H&Z had made correct observations and predictions for most of the types. However, I found data that were not compatible with H&Z on five occations.
- My informants did not accept a finite verb to the left of the adverb ikki.
- They do not accept lon-distance binding
- They do not accept a PP complement for the benefactive argument of give-verbs
- They do not accept a prenominal genitive: *Ólavsa bilur
- They did not like nominative subjects here: Honum varð heilsað, *Hann varð heilsaður
For those interested, please go to the Document Chest, where I put a summary of the results that I found.
– Janne
From Philipp Conzett:
 |
| Philipp Conzett eliciting Faroese gender assignement from a Fuglafjørður informant. (Photo: Maja Sojtaric, Labyrint/UiT) |
This NLVN and NORMS seminar was my first experience as a field worker, so it was good training. During the visit I did some investigation into the gender assignment system of Modern Faroese. Since Hjalmar Petersen has been working on this issue, too, we joined forces to create a questionnaire.
1. Gender assignment to non-existing nouns
The first task in our questionnaire was to assign gender to some non-existing nouns. In his PhD thesis, Hjalmar puts forward some semantic gender assignment rules. Three of them are given in (1):
(1) Semantic gender assignment rules in Faroese (Petersen 2008:140-144):
a. Trees are masculine or feminine
b. Fish are masculine or feminine
c. Birds are masculine or feminine
So Hjalmar suggested to test these rules with some non-existing nouns.
In the interview I asked the informants to imagine that there had been found a new kind of tree, fish, and bird, respectively. Hjalmar and I had prepared some sentences where we used the same made up noun stem occurring with different genders, and within the same gender also with different inflection classes. So, I then first show them a drawing of an imaginary new fish, tree ord bird and then asked them if they could say those sentences about this new fish, tree or bird. So, the sample for the fish looks like the one in (2):
(2) Non-existing nouns, Sample 1: [Drawing of an imaginary fish]
I have only had the time to see at the answers from two places, Fuglafjørður and Tvøroyri. I interviewed six informants at each place. The examples in (e) were really confusing for the informants, so I dropped the sentences after the first three or four informants. The reason why they got confused is probably that this inflection class is only used for abstract nouns like elli meaning ‘age’.
I haven’t investigated yet if there is any dialectal variation between two places, but in (3) the numbers from all twelve informants from both places are summarised. As you can see, for the new fish, the masculine in -ur was the most preferred one, that is the form ein pivur. Second most natural sounded the feminine noun ein piva to the informants. For the new tree, the feminine noun ein seipa was most preferable, but ein seipur would do almost equally well. As for the new bird, the strong masculine ein palur got the top rate, but both the weak masculine ein pali and the weak feminine ein pala scored almost equally high.
(3) Preliminary analysis of Sample 1-3:
2. Gender variation
In the second section of the questionnaire we wanted to find investigate some existing Faroese nouns that are reported to have different genders in different areas. I got a list of possible candidates from the local work shop organisers. I included eleven of these nouns in my data set. They are all listed by their stems in (4).
(4) Nouns with reported gender variation in Faroese:
a. Inherited (simplex) words: verkstað- , skúr-, ramm-, skóg-, ítrótt-
b. Established loan words: køk-, fíl-
c. Newer loan words: café-, trend-
d. Derivations: betring, hending
The data sample includes different types of words. As you see, there are some inherited nouns in (a), some established loan words in (b), some newer loan words in (c), and some derivations in (d).
In the two dialects I have looked at so far, I couldn’t find any clear pattern of variation, apart from the noun køk/køkur, which Zakaris already mentioned in his talk. In (5) I have summarised the figures from Fuglafjørður and Tvøroyri. In Fuglafjørður only one informant answered that he maybe could use the feminine noun køk, whereas in Tvøroyri it seems that most informants could use both the feminine køk as well as the masculine køkur.
(5) Gender variation in then noun køk-:
My impression from the other places we visited, is that there is some more dialectal gender variation in the other dialects, but I haven’t seen at the data yet.
In the interviews I noticed that the noun trend/trendur probably was the leas established word in my investigation. Many people were unsure about what gender they should use with this noun. This is also reflected by the figures for Fuglafjørður and Tvøroyri. As you can see in (6), the all three genders are attested for this noun. There were also several informants that accepted all three genders. This would be in correspondence with what has been reported from other languages, namely that there can be some transitional period where the gender of some loan words vacillate.
(6) Gender variation in the noun trend-:
Further analysis will hopefully soon be published by Hjalmar and me.
– Philipp
From Saija Tamminen-Parre:
 |
| Saija Tamminen gathering information about pragmatic particles in Fuglafjørður. (Photo: Øystein A. Vangsnes) |
In my research on Faroese I looked at pragmatic particles, and with the data I collected, I plan to compare the use of pragmatic particles in Icelandic and Swedish to the use of similar particles in Faroese. I want to see if such particles in Faroese are used in a similar or different way from those found in the former languages. I would also like to look at the use of the same pragmatic particles in free speech (e.g. as found in conversations in Faroese taped by other NORMS researchers), and compare those results with my own questionnaire results, to check how people think they use pragmatic particle and how these particles are actually used in free speech. I interviewed people of different age, gender and from different places, in order to check if there is some variation depending on/according to these variables.
– Saija
From Höskuldur Thráinsson:
During the field work in the Faroes I was mainly looking at the following:
• verb placement in embedded clauses of various types
• expletive constructions of various kinds and possible subject positions (positions fot the associate of the expletive) in these
• sg./pl. distinction in the past tense of regular verbs
I tried to break the sessions up by having three different tasks. In the first part I would ask the speakers for “absolute judgments” using a three point scale with the following definitions:
“Yes” = A good sentence. I could easily say this.
“?” = A doubtful sentence. I could hardly say this.
“No” = An impossible sentence. I could not say this.
 |
| Höskuldur Thráinsson running his questionnaire on a Fuglafjørður informant. (Photo: Øystein A. Vangsnes) |
This is the scale that we ended up using in our large scale overviews in Iceland and I explained it to each subject before we began the sessions. I had a simple, straightforward natural sentence as the first example and then an impossible variant of it as a second example to kind og “set the stage” and boost the speakers’ self-confidence. I did not, however, include any context sentences and regretted it. It would have been quite useful and important for some of the examples I had, especially the expletive constructions, which seem to be especially sensitive to context.
After this first part I then had a few pictures that I used to try to elicit present and past tense forms of verbs (mostly regular ones). This worked quite well for most of the speakers (although one of them constantly refused to give me past tense forms, which I was trying to elicit by using the adverbial “í gjár” ‘yesterday’: The speaker (correctly) pointed out that the pictures did not show what had happened ‘yesterday’). I found it useful to break up the sessions in this fashion – it tended to make the speakers more relaxed and also gave me the opportunity to ask about different things, e.g. different lexical items.
In the final part of the sessions I would then ask the speakers to compare two variants that mostly had to do with different word orders, both verb placement in embedded clauses, subject (expl. associate) placement in expletive constructions with one stylistic fronting example thrown in for good measure. I asked the subjects to pick the variant that they thought sounded better or tell me if they thought both (or all – I had three in one case and four in one instance) were natural. If they picked one over the other(s), I would ask them if the dispreferred one(s) were completely impossible or less natural or would be used in particular contexts. I had the subjects read the sentences aloud before they judged them and this turned out to be useful since some subjects (not many, though) would occasionally misread the examples. I taped all of the sessions so I can go back and check things, but some of the recordings are rather useless since several speakers spoke rather softly and there was often considerable noise in the rooms.
 |
| Faroese tiny village and fog. (Photo: Tania Strahan) |
I tried not to have a huge number of sentences and the sessions would usually last less than 20 minutes (often around 15 or so). Ásgrímur and I tried to interview many of the same subjects since we were interested in some of the same things and had coordinated our questionnaires to some extent. I will put the tabulated results in the “dokumentkista” when I get around to it, but my first impression was that there was considerable individual variation and not clear differences between the places we went to – except perhaps that the speakers in Sandoy were less likely to “move their verbs” than other speakers, despite the fact that the most famous verb mover in the Faroes, the author Heðin Brú, comes from there. It also seemed to me that the type of adverb played a role, as it does in Northern Norwegian and perhaps also in Älvdalen, although clause type may also be relevant (verb movement in bridge verb complements generally well received but verb movement across negation in relative clauses rejected by most of the informants I talked to).
In general, I thought the whole thing worked extremely well – the local organizers had managed to get amazingly large groups of speakers and the speakers were very cooperative, positive and interested. Comments revealing worries about prescriptive norms seemed really exceptional.
– Höski
From Øystein A. Vangsnes:
 |
| Øystein A. Vangsnes and an informant at the school in Fuglafjørður. (Photo: Maja Sojtaric, Labyrint/UiT) |
I took part in four of the altogether six sessions during the fieldwork, and Gunnar Hrafn tested out my questionnaire at the remaining two places, so I have data from all six locations and from 43 different individuals in total. My questionnaire focused on issues relating to wh-syntax, mainly: (i) degree questions, (ii) wh-nominals, (iii) exclamatives, (iv) binominals, and (v) wh-extraction. I also tried to test (vi) whether the rising tone in questions is common to all Faroese. In total I had well over 60 example sentences, but not all were used on all informants. This time I didn't use a scale – I read all of the examples out aloud and then asked the informants whether they found the examples acceptable or not. If they were uncertain or had specific comments I made a note of it. It took about 10 minutes to run through the questionnaire with each informant. The results are available in the form of a spreadsheet in the Document Chest.
Some of what follows repeats stuff from the Wednesday 13 August entry and from a comment to Gunnar's Thursday 14 August entry.
(i) Degree questions:
Like Swedish and English, Faroese uses the same wh-word (hvussu) in degree and manner questions. I wanted to check whether Faroese speakers could also use hvat ‘what’: 13 of the informants actually accepted the example Hvat gamal er hann? ‘How old is he?’, three of which however commented that only young people say this. A few others noted that they hear people say this or were uncertain. However, none of these informants accepted examples like ‘What big was the fish?’ or ‘What many cars have you?’, so there might be something special about the particular question for age. Interestingly, four Sandoy and one Vágar informant accepted the Icelandic word order ‘What are you old?’, two of which rejected ‘What old are you?’ in the first place. At this point Gunnar Hrafn had taken over the questionnaire, which was good, because I had unfortunately stopped asking about this example after getting only negative evidence up north in Fuglafjørður and Klaksvík (and in Tórshavn). All in all, it seems that the use of hvat in degree questions is marginal, but it might be worth investigating further.
A related issue, I thought, concerns hvussu vs. hvat with the verb eita ‘be-called’ (cf. German heißen). Standard Faroese has Hvussu eitir tú?, which everybody accepts. 17 informants also accpted Hvat eitir tú?, which is in line with most other Scandinavian varieties. However, there is no obvious correlation between those who accept ‘what’ in ‘How old is he?’ and those who accept ‘What is he called?’: The overlap is 8/13.
 |
| Faroese fjord. (Photo: Tania Strahan) |
(ii) Wh-nominals:
On this issue there is a whole lot to say – in fact, too much for this report. One interesting fact about Faroese is that it is the only variety of Scandinavian which, as far as I know, uses inflected ‘what’ as a determiner (as in Old Norse). An example would be Hvønn bil eigur tú? ‘What-ACC car own you?’. At the same time the language has acquired a "what for (a)" determiner. For some speakers at least it seems that the former triggers a token reading whereas the latter is neutral wrt. the type/token distinction. Icelandic neither uses inflected ‘what’ anymore, nor has acquired a ‘what for’ determiner (but has the uninflected hvaða for token queries and overt kind expression in kind queries. For the ‘what for (a)’ determiner I tested out various variants and most people accepted both the presence and absence of ‘a/some’ after ‘for’. All in all the judgments appear quite uniform accross the informants. One point of variation, however, was to what extent bare ‘what’ also could be used determinatively: 7 informants accepted the example Hvat bilur er tín?’, and 4 of these came from Suðuroy. As a final remark: I found no signs of the adnominal use of ("manner") "how" which is found in colloquial Icelandic and many Mainland Scandinavian varieties: no speaker accepted Hvussu bil hevur tú? ‘How car do you have?’.
(iii) Exclamatives:
Something completely new to me was the way in which Faroese exclamatives are standardly construed, namely by an initial sum, i.e. the relative/comparative etc. complementizer corresponding to som/sem in Mainland Scandinavian and Icelandic. So for an exclamative where the scale is provided by a stative predicate (as in "How tall he's become!"), all informants agreed that a sentence like Sum hann er vorðin stórur! was fine and for an exclamative where the measure is DP internal (as in “What a nice car he has!”) there was full concensus about the sentence Sum hann hevur fínan bil!. Icelandic apparently has a similar option using Það sem... (‘That which/SOM...’), but other “Icelandic” ways of forming exclamatives (such as a fronted intensifier or wh-degree element) were not readily accepted. Another Faroese way of construing exclamatives, which was accepted across the board, was the particular embedded structure exemplified by Tað er ótrúlegt hann er vorðin stórur!, i.e. word by word “It is incredible he is become tall!", in other words without any overt degree word.
I also checked the so-exclamative types well-known from Mainland Scandinavian, and about half of the informants accepted this, i.e. examples like So stórur hann er blivin! and So fínan bil hann hevur!. In any event it seem quite clear that the sum-type exclamative is the general one. Afterwards I have established that there are reasons to believe that it is sum qua comparative, and not relative, complementizer that is involved in the exclamatives as it cannot be replaced by the exclusively relative complementizer ið (which is used with subject relatives and clefts). This squares with the general insight that there is an affinity between exclamatives and comparatives.
(iv) Binominal each:
As far as binominal constructions are concerned I have little extraordinary to report. As in the other Scandinavian varieties such constructions involve the quantifier hvør, which in Faroese, however, still is identical to the wh-pronoun ‘who’ (and the determiner ‘which’), cf. above. Just like in Mainland Scandinavian we find to different structures, one where the quantifier is at the left edge of the second DP and followed by a reflexive pronoun (Vit hava fingið hvør sína bók ’We have got each REFL book’) and another where the quantifier is at the right edge of the DP akin to the English construction (Vit hava fingið eina bók hvør. ‘We have got one book each.’) The quantifier agrees with the first DP in case, i.e. the one that binds the second one, whereas the rest of the second DP carries a different case. One interesting option in the latter construction is to replace the quantifier by the PP í part ‘in part’: I haven't heard about that from any other language.
(v) wh-extraction:
One issue concerning wh-extraction in Faroese is that quite a few of the speakers consulted accept insertion of the relative complementizer ið when the fronted wh-constituent is the subject of the embedded clause. Such insertion is not accepted with wh-object, and this then parallels data well-known from Norwegian dialects concerning som-insertion. The specific figures concerning ið-insertion are that 16 of the 43 informants accepts this and one says "maybe": only 5 accepts sum instead of ið (and a sixth says "maybe"). As for the geographical distribution none of the informants consulted in Vágar and Tórshavn accepted ið-insertion; in Suðuroy 5 of 8 accepted it, in Klaksvík 5 of 10 and in Fuglafjørður 4 of 8. Maybe a tentative conclusion could be that this is a trait of "peripheral Faroese”?
I have also since learned that comparable der-insertion is readily acceptable in colloquial Danish, and given that this is on the ScanDiaSyn-questionnaire (at least for Norwegian) there are good chances that we'll have a good picture of the geographical distribution of this phenomenon pretty soon.
Before leaving the islands and leaving the questionnaire to Gunnar Hrafn, there were all informants showed a *that-trace effect by rejecting the sentence Hvør heldur tú at __ hefur gørt tað? ‘Who do you think that has done it?’. In Sandoy Gunnar encountered two informants who accepted it and a third one who were uncertain. All of the other 40 informants reject the sentence, so I think the picture is pretty clear for Faroese on this issue.
(vi) Other:
For the fun of it I tried, at the beginning, to see if anyone would accept main clause wh-questions without Verb Second, but that soon turned out to be a waste of time, so I stopped. I also tried to elicit data concerning the putatively characteristic rising tone in Faroese questions, but I'm very uncertain about the results that I got. The way I did it was to ask the informants how they would ask me whether I have been to Iceland, and then upon producing the question I would pay attention to the sentence final prosody. Surprisingly many would in fact produce the yes/no-question without a rising tone, and many of them would furthermore reject that option. But I'm not sure about the validity of the test, so all I would like to conclude at the moment is that this should be further investigated, and presumably natural conversation would be a more ideal basis for doing it. And then it would not surprise me if the rising tone is subject to geographical variation, with havnarmál being the center for the phenomenon.
– Øystein
|
|
|
