Saturday, August 13, 2022

Real Language Examples: Traces of the Old Swedish Case System in Modern Swedish

övermåttan - exceedingly, over the measure(acc)
an example that well describes this essay.

Traces of the Old Swedish Case System in Modern Swedish

This article is mainly written as a potential source of inspiration for conlangers, especially those whose focus lie with naturalistic constructed languages. This comes with a caveat emptor: claims here may be incorrect. I have gone to some effort to verify stuff, but mistakes may have slipped by. With some additional effort on the part of the reader, it could perhaps suffice as a scholarly source. However, paradigms for Old Swedish, Old Norse and Old English are largely taken from wiktionary. Information about the extant or recently extinct case systems in dialects are hard to find reliable sources for.

I know of no previous list of all remnants of the case system in Swedish, and writing this has been kind of annoying. Every time I've thought I've had it entirely covered, an expression I haven't used for years rears its head, some expression in the newspaper catches my attention, I say something and realize one second later that it's yet another example, I hear someone speaking dialect come up with yet another example, ...

Swedish, much like English, has lost the old Germanic case system rather thoroughly. Both languages retain one case distinction - nominative vs. genitive, where the nominative mostly has taken over all the roles of the dative and accusative, but also partially some roles that the genitive previously has had. This is an interesting opposite to modern German (and Älvdalska) - where the genitive was lost but the others kept.

Unlike English, however, Swedish does keep a fair few traces of the dative and accusative, as well as a somewhat more robust trace of the genitive's previously wider function. Whether to consider the genitive a case or not in English and Swedish is tricky, but there may be slightly stronger reasons for doing so in Swedish.

Let's first quickly review the case system of modern Swedish, and contrast it with the case system of Old Swedish. The developments should be familiar to anyone with any knowledge of English historical linguistics. Several of my statements below are from this source.

Old Swedish and Case 

Old Swedish had a typical Germanic four case system: nominative, dative, accusative, genitive. The case system was never fully distinct, i.e. some case syncretism has always existed, at least for some cases in some genders. Oftentimes, only adjectives and articles would mark for case (much like modern German!), and these markers seem to have survived for longer than markers on the noun itself. The syncretism naturally also had less of an impact in any noun phrase were multiple words had different patterns of marking present on them, as the actual intended case in some sense could be triangulated from the noun and adjective markers.

    Morphology 

A complication that is shared with German, is that case morphology on adjectives varies by certain parameters (strong vs. weak, which mainly correlates with indefinite vs. definite). In addition, case morphology for definite nouns differed from indefinite nouns, and you get a wild variety of declensions with their own peculiarities. Definite nouns and indefinite adjectives carried the main load of case morphology.
 
There is an appendix at the end detailing the Old Swedish nominal and adjectival morphology. A detailed knowledge of it is not necessary to understand the remainder of this article. Besides, throughout the decline of the case system, it is fairly likely all manner of modified versions existed for short time-spans or regionally. Besides, we do also find changes between Old Norse and Old Swedish, where nouns change conjugation or whole conjugations change a bit. Largely, the system was preserved, but details did change.

The adjective morphology was in some sense simpler than the nominal morphology - almost all adjectives behave the same way all the time, using the same morphemes, etc.

Syntax

Much like in pretty much all other Indo-European languages ever, each preposition took some particular case, and the cases also sometimes affected the meaning of the preposition. It is hard (or expensive) to come by information on this. We should probably talk of adpositions rather than prepositions in Old Swedish and Old Norse, since they were much more frequently used as postpositions than they are today. Today, this nearly exclusively survives in a few fixed expressions.

Genitive expressions were originally head-initial, but this probably changed during the course of Old Swedish. Remnants of the postposed possessor still occurs with possessive pronouns even in the written language, but is increasingly considered dialectal. Swedes in Sweden really don't like anything actually conservative in the written language ever, and thus this will probably go extinct soon enough.

In properly old Norse, we find quirky case. A possible trace of this, somewhat similar to English 'methinks', is 'det ante mig'. Ana = suspect, perceive, but somewhat at the edge of perception. In modern Swedish, this structure would be a slightly malformed "It suspected me", but signifying 'I knew it!' (In modern Swedish, 'ana' is 'anade' in the past tense, not 'ante'.)

We find in Old Swedish that the genitive can be used as an adverbial case, i.e. it is not restricted to adnominal and adpositional usages. Many modern retentions of the old genitive in fact reflect such usages, and this is often evident with a variety of adverbs pertaining to time or manner.

Modern Swedish and Case

Morphology

Modern Swedish has nominative vs. oblique solely in the pronouns - with possessives formed either as genitives or as possessive adjectives. This is somewhat similar to English. The noun is - like in English - simpler, having the nominative and the genitive, which is nearly invariably marked by -s*. The pronominal oblique is used after prepositions, before postpositions and as objects of verbs.

Nouns generally, but not always, correspond to the historical nominative form. A few exceptions where historical obliques won out exist. Modern nouns have one case-like marker:

-s is a clitic that goes on the last word of the NP that is the possessor. Thus "kungens halspastiller" ('the throat lozenges of the king'), kungen av Danmarks halspastiller ('the king of Denmark's throat lozenges'). A few conservative speakers and writers may prefer an older order where "the king's of Denmark throat lozenges" hold.**

The possessive suffix goes on the end of any inflected form of the noun, in an invariant form. It may not go on a stem if the stem deviates from the nominative. Nouns that already end in -s have a phonological rule that deletes the suffix. (I.e. we don't do any of that silly -s's-stuff.)

* The sole exceptions are marked by zero, and are nouns whose stem end in -s. The -s suffix can be attached after definite suffixes and numeral suffixes as well, giving e.g. sg: hus / huset-s, pl: hus / husen-s, sg: bil-s / bilen-s, pl: bilar-s, bilarna-s.

** (The canonical example is 'kungens av Danmark bröstkarameller', i.e. 'chest candies', but this just sounds weird to English speakers. Apparently they were aniseed-flavoured with some beetroot for colour.)

As for -s going on the final word of an NP, this means that something like

mannen som kom ins skor gnisslar
the man who came in's shoes squeak

is entirely acceptable in colloquial Swedish. SAG claim that some words, however, do not tolerate -s on it even in such positions, and include some personal pronouns in that class. Rewording seems the only reasonable solution in such cases - omitting the -s is not permissible. 

Syntax

In some varieties of Northern Swedish, the definite form sometimes is used in a way that seems as though it might be influenced by the Finnish accusative and partitive, or that in some other manner strongly deviates from the way it is done in Swedish. My dialect has some traces of this, but as it is strongly restricted by the lexicon (only a semi-closed class of nouns do this) and by semantics, rather than by syntax and semantics, I hesitate to call it a case as far as my dialect goes. It may be that it is more case-like in dialects further north. However, even then you have to account for how the definite seems to operate in any syntactic position even before the -s genitive.

The -s genitive is nowadays used besides adnominal and complemental possessives, as a stand-in for nouns (Peter's book was boring but Elof's was interesting), and with some restrictions with some prepositions.


The Case System in the Modern Swedish Pronouns

Morphology


I you (sg) he she it itcommon reflexive we you (pl) they
nom jag
du       
han
hon
det
den
--
vi
ni
de
acc mig
dig
honom
henne
det
den
sig
oss
er
dem
poss
min
mitt
mina

din
ditt
dina
hans
hennes
dess
dess

sin
sitt
sina

vår
vårt
våra

er
ert
era
deras

 

Itcommon is a gender that emerged in the 18th century, into which non-humans nouns that previously were masculines and feminines were moved. In modern Swedish, it is called 'utrum reale', sometimes just utrum or reale. Conservative dialects often retain the three-gender system.

As you might see, there are large similarities with English. Unlike English, however, the possessive pronouns for my, your, the reflexive, our and your (pl) have gender congruence. For this reason they're not considered case forms, but some form of possessive adjectives. NB: terminology on this varies from author to author.

Swedish also lacks the my/mine-distinction. "Mine" in some sense has some case-like properties that differentiate it from "my", thus presenting English with a two-genitive system. (I am here thinking of 'of mine' and such).

It, and itcommon sometimes get the genitive form dens, dets, but these are considered substandard or wrong. "Hons" as a possessive of "hon" does occur in some speakers idiolects, and of course anything like "jags", "hons", "hans", "vårts", etc can occur if you create a relative subclause for a genitive noun and force "jag", "han", "hon", etc to be the final word of the subclause, e.g. something like

the guy who talked faster than I's book
snubben som talade snabbare än jags bok

However, these will probably feel somewhat off to a significant number of speakers. (My personal view is that -s in Swedish and -'s in English are postpositions that go after an NP. These are ok for me in spoken Swedish, but not in written Swedish.)

Swedish has had a who/whom-distinction (ho/vem), but the oblique form has won out and hardly anyone even recognizes "ho" anymore. The last time it was used in print was in the 1940s in a hymn book for certain hymns from the early 19th century. The third person plural pronoun is nowadays in colloquial Swedish mostly "dom", with the de/dem-distinction still surviving in Finland. I am unaware if any colloquial speakers have any kind of three-way distinction between de/dem/dom. (Imagine how cool it would be to discover some Swedes having a tripartite alignment with 'dom' for intransitives or something like that!)

Historically, "han" was both the nominative and accusative. "Honom" was exclusively dative. Many dialects and almost all colloquial Swedish retain "han" for the accusative (and dative as well). I am unaware whether any speakers of any colloquial variety retain "honom" exclusively for the dative - this would be interesting. Some dialects have further aånalogized the masculine situation and merged "henne" into "hon", sometimes with even greater loss of oblique forms of the pronouns (tho' the obliques may be retained as emphatic forms; here, my dialect is weird, having emphatic nominatives and emphatic accusatives, yet being on the way to turning the accusative and emphatic accusative into optional adornments).

Finally, 'sin/sitt/sina', the possessive reflexives, could maybe be argued to be oblique forms, as they can not stand as possessors of subjects. They can, however, stand as possessors of nominative complements, so this argument would be somewhat weak.

Examples by type

As mentioned, examples of the case system survive in fossilized phrases in Swedish. This strives to be an exhaustive list of such examples.Each type of example probably also has dialectal examples en masse in a variety of dialects. Because my dialect is a subtype of middle Ostrobothnian, this will be overrepresented.
Omitting the bolded letters should give the nominative form, otherwise, I will provide the modern nominative in parentheses. If no properly corresponding nominative exists, it is marked by a dagger.

1. Fossilized expressions with case morphemes

hemma (from 'hem' - home) historical genitive, signifies 'at home'. Hem, which is either the noun used when it is the subject or object, also is used as a directional adverb (i.e. towards home), due to this also being the historical accusative.

i ljusan låga - in bright flame - historically 'låge' was nominative, but låga has replaced it
mitt på ljusan dag - in the middle of bright day
sova i godan ro - sleep in good calm
i gladan håg - "in a glad mind", 'håg' is also an archaic noun by now.

taga till orda - to commence a speech, 'take to word'.

till väga - by (some) means, by way of ... mostly in the expression 'hur skall vi gå till väga' (how shall we go about something/by what means shall we proceed) and similar.
av daga - 'off/from day' - in the expression 'taga av daga' - to kill. ('to take away from (the) day')

i sinom tid - in its time

Basically often used to mean "after some time".

allom bekant - familiar to all
androm till varnagel - (andra) for others as a warning.

"Androm till varnagel" has probably been extinct, but saw a slight renaissance after being used in the title of a book.

ana argan list - suspect angry cunning (here, originally 'arg' probably was rather meant as 'evil')

gammal i gårde - "old in the farm", i.e. experienced

man ur huse - "man out of house"

Mostly in the expression 'gå man ur huse', which expresses something so popular or controversial that people come out in the streets to protest or acquire the thing.

dra sina färde - pull one's voyage - to leave.

fara å färde - danger on voyage, 'danger skulking about'

med vett och vilje (vilja) - with knowing and will, 'knowingly and willingly'

till viljes (vilja) - to will, i.e. "göra någon till viljes" - do someone to will, "to appease someone"

se i syne (sometimes generalized, giving höra i höre) - see in sight, but figuratively 'hallucinate a sight'.

se i andanom (ande, -n) - to see/perceive "in spirit", apparently sometimes used for 'in one's imagination'.

med råge - (råga, supplanted by råge). 'with excess'.

2. Genitive usage in expressions of time and with prepositions

tids nog - 'soon enough'

Essentially "time.GEN enough".

There are a number of expressions where the genitive or the nominative are used, with a slight difference in meaning. Generally, the genitive will signify the most recent such, whereas the nominative will signify the upcoming one. Some of the forms are morphologically slightly exceptional - generally keeping an extra vowel that would be lost in the possessive genitive.

i morse †- this morning (unclear morphological relation to morgon - morning)

i kvälls - last evening
i lördags - last saturday (works with any weekday)
i somras (sommar) - last summer
i höstas - last autumn
i våras - last spring
i vintras (vinter) - last winter
i julas - last christmas
i påskas - last easter
?i pingstas - ?last pentecost
?i midsomras - ?last midsummer
*i fastlagstisdags - last lent tuesday
*i januaris - last january

I think I might have encountered reanalyzed use of these, i.e. people using "somras" as a shorthand for the nominative "(senaste) sommaren" when signifying the most recent summer, thus enabling things like "höstas var trög, våren blir säkert bättre" ('last autumn was slow, the spring certainly will be better'). My impression is that this is very marginal.

My dialect has this strange form for yesterday morning:

i gåron måronen (c.f. standard Swedish 'i går morse')

I am inclined to believe -on is a retention of some case marker, but sound changes seem to have garbled it a bit, and it thus becomes difficult to figure out what exact case it might have been. Also, it's been a bit difficult to find morphology for "igår" (yesterday), since this only seems to survive as a set phrase in all the cognate languages.

I am not sure how to feel about the fact that the above examples all have -st instead of -s in my dialect, whereas possessive genitives and the next examples all have -s in both the standard language and my dialect. This might suggest it originally was some other type of derivative morphology in the above examples.

Several somewhat established "places" and methods of movement and miscellanous fixed phrases still take the genitive after 'till':

till sjöss - by sea
till havs - by sea
till lands - on land,
till skogs - in the woods
till fots - by foot,
till båts - by boat,
till hjuls - by wheel
till hands - available ('to hand')
till råds - for an advice ('to advice's')
till sängs (in bed, into bed)
till låns - for a loan, as a loan, borrowed
till livs - for nourishment (not 'alive')
(slå någon) till döds - (beat someone) to death
till nöds - "to need", often used to express 'as a last way out, as a sufficient but not really good substitute'; nöd nowadays rather means 'severe, immediate need or danger'.
(alla) till lags - 'vara alla till lags' signifies 'satisfying everyone's demands'.
(illa/väl) till mods - being in a bad/good mood. 'Mod' nowadays means 'courage', not 'mood'.
till  bords - aboard
till äventyrs - to adventure, more specifically 'unexpectedly'
tillbaks / tillbaka - back, in return.
till ords - in word(s).
(ta) till bruks - 'ta till bruks' signifies putting something to use.
till gårds - in(to) the yard.
till undantags - for an exception. (undan = away, tag = take; 'for awaytake')
till vardags - for everyday use. "var" is here from the same root as "varje", i.e. 'each'.
till buds - on offer. ('bud' is the noun cognate to 'bid', 'bjuda' would be the verbal form)

"Till råds" is used in somewhat different ways in different expressions, mainly expressing the notion of 'consulting someone': ta till råds ('take for an advice'), fråga till råds ('ask for an advice'), etc.

?till stads - in (the) town (dialectal?), often also till stass.
?till bys - in the village

Literally "in (the) village", but idiomatically always used in the sense of 'being someone's guest' (dialectal?). "Vi var till bys" - 'we were on a social call, we were visiting someone'.

+till vagns - by wagon
?till peds - by bicycle
?till cykels - by bicycle,
?till sparks - by kicksled

I believe "till vagns" is extinct in the wild. The last two I'd use in dialect, but not in the standard language. I am convinced 'till peds' exists in several dialects. Other prepositions, mainly nowadays turned into compounds:

utomlands - abroad
utom laga tid/laga bot/... - beyond legally mandated time/legal fines/...
inombords - on the inside
utombords - on the outside
med ens - at once, "immediately", literally "with one's"
motvalls - adjective signifying 'contrarian', literally "against the embankment'

medsols - "with sun's", i.e. clockwise
motsols - "against sun's", i.e. counterclockwise


Further, the words corresponding to 'hundreds of', 'dozens of', etc is formed using the genitive on the numeral no matter what syntactical role the NP has. However, the numeral for these is formed with 'tal' (number, cognate with 'tell') as suffix, which carries the genitive marker. The noun itself is in the nominative, unless it is a possessor.

tiotals - 'tens of'
dussintals - 'dozens of'
hundratals - 'hundreds of'
tusentals - 'thousands of'
tiotusentals ...
hundratusentals ...
miljontals ...

Other obliques as object or after prepositions:

i min ägo († äga) - 'in my possession'
i hennes närvaro († närvara)- 'in her presence'

Närvaro has been lexicalized as a nominative by now, so e.g. "närvaro är obligatorisk": presence is mandatory.

komma i delo - 'to get into a strong disagreement'

Del signifies 'part', but here, 'delo' simply is 'apart' or somesuch.

sätta å sido (sida) - put aside. "Å" is a semi-fossilized preposition in Swedish.
utan återvändo († återvända) - without return
skämt åsido : "joke(s) aside", a common phrase after saying a joke and I don't even get what the point of this phrase really is.

till salu († sala) - for sale

till fullo - fully ('to full')

till yttermera visso - to outermore wit ("in addition", "also")
till godo - for benefit, for gains, to (some kind of figurative or literal) credit
av ondo - of evil
i lönndom (†, sorta) - in secret, no corresponding nominativ.

till spillo - (something goes) to waste, spill is not very commonly used, but is basically the nominative form.
i så måtto - to such an extent, to the extent, to what extent ('in so measure')
över måttan - exceedingly, 'over the measure'
fatta posto - to occupy a position (c.f. the second meaning of 'take post')
under någons domvärjo (-värja) - beneath someone's sword of judgement, "under someone's jurisdiction".

åtminstone 'at least'

Some of these also appear in compound verbs: åsidosätta (put aside), tillgodogöra (to compensate, literally "to make for a benefit"), saluföra (literally "sale-lead", to market, to sell, to mongle).

3. Old genitive forms in certain types of expressions

Towns and villages whose names end in a vowel in some contexts get a suffixless genitive:

Åbo Hovrätt - Turku court of appeals
Vasa stadshus - the town hall of Vasa ('Vasa town house')
Visby mur - wall of Visby (Visby wall)

In other contexts, the -s is expected, e.g. "Åbos nyaste misstag" - the newest mistake of Turku.

The plural genitive marker -a appears on "lag" (law) in some expressions:

i laga ordning - in lawful order
laga förfall - lawfully valid excuse
i laga kraft, i laga tid, med laga stöd, med laga rätt - in lawful validity, in lawful time, with legal support, with legal right.

4. Ackusative (or oblique) forms that have become their own lexemes

Besides the already mentioned närvaro and ägo, another -varo word qualifies: tillvaro ('existence', especially in the sense of focusing on the quality of said existence).

Other examples include:

ande / anda

Ande signifies the spirit of a being. Anda is rather 'the spirit of a moment, movement, time'.

fura / furu

Fura is a specimen of pine, furu is pine, the material.

Apparently, several types of tree and wood have, at least dialectally, such pairs, e.g. björk/björke (birch), rönn/rönne (rowan), asp/aspe (aspen), en/ene (juniper), ek/eke (oak), sälg/sälge (sallow), bok/boke (beech), pil/pile (willow)  where the -e-variant is the wood material, and the e-less-variant is the actual tree.

dag / dager

In Old Swedish, 'day' was nom: dagher, acc: dagh, dat: {dæghi, daghi}), gen: daghs. Dag nowadays means 'day', dager means 'daylight'.

mosse / mossa

These are cognate to English 'moss'. A mosse is a type of swamp, mossa is moss.

grädde / grädda

Cream, cream of the crop/the social elite.

ull / ylle (wool)

With these two, it seems to me that the overlap is a bit greater than with the previous examples: while on the sheep, it is only ull, but after that point, both are used, but then for compounds like "ylletröja" (wool sweater) only "ylle-" is possible. "Den är av ylle" (it is (made) from wool) is however probably also mandatorily 'ylle' rather than 'ull'.
 

Although this is not an example of case, a similar development has happened with the distinction between trä ('wood', as in the material) / träd ('tree', as in the plants) , where a former definite form ('träð') has become a new indefinite singular form with a slight change in meaning, and we thus have the historically 'double definite' "trädet" (the tree), vs. "trä(e)t" (the wood).

ände / ända

Visually very similar to the spirit/metaphorical spirit-pair, this pair is at the other end of human experience: end, ass. "Ända" sometimes does appear in expressions such as "till ända" ('at end', i.e. finished), and "ände" sometimes does signify "ass", so this pair is somewhat ~statistical. Ända is thus statistically more likely to signify the ass end, whereas ände is statistically more likely to signify other ends.

 

5. Datives and obliques in place names

Topononomy will be overrepresented in prepositional phrases, and it's likely for oblique forms to feed back into the nominative for this reason. Several places have retained the plural dative - as their only form, replacing the nominative - for this reason, here present either as -om or -um:

Sundom - strait (sund is cognate to 'sound')
Husom - house (hus is an obvious cognate)
Pörtom - pörte is a type of house with a smoke hatch rather than a chimney
Bodom - hut ('bod')
Husum - house, again. -um and -om might have been in free variation
Kvänum - seems disputed, probably rather *Kvädnhem, so the qued home, from the name of a river.
Lerum - clay? (Unlikely?)
Vattjom - ?
Skedom - spoon?
Torrom - ?
?Salom - halls?
Skadom - damages (?)
Lökom - onion?
Mjällom - in modern Swedish, mjäll is dandruff. I doubt this is the correct etymology.
Östmarkum - east land ('mark')
Västmarkum - west land ('mark')
Arlom - ?

The singular dative -e may also have survived, but since -e can appear in nominatives as well (an old masculine marker), this is harder to spot on a map. I imagine Tjärne, Höje, and several -inge may have such an origin, but it is hard to weed out false positives with this.

The singular oblique -a also appears:

Grundsunda - shallow strait.gen
Kasabacka - pile.gen hill.gen
Bergsboda - mountain.gen hut.gen
Ytterboda - outer hut.gen
Tobacka - ? hill.gen
Alberga - Alder mountain.gen
Vänoxa - friendly ox.gen (?) - I bet there's some other etymology there
Kårkulla - ? hill.gen
Torsholma - Thor's islet.gen
Björnholma - Bear islet.gen
Västansunda - west.obl strait.gen
Aspa - aspen.gen?
Fjälla? - mountain.gen
Stavsudda - staff.gen ness
Haga - hedge or meadow (from hage). Cognate to The Hague.


-u/-o are also an oblique form that may be present in at least a few place names:
Korpo, Nagu, Sagu, Sorpo?

It is conceivable that some toponyms ending in -an are historical accusative adjectives, but since -an also is a nominalizer in more modern Swedish, this is hard to be sure of. Whether Långu outside of Stockholm is a dialectally altered "Långö" (long island), or an oblique form is hard to tell. I am not aware of many islands whose names are just adjectives, (whereas such names seem fairly common for lakes and rivers).

Pedersöre, Örebro, Öresund may contain a dative of "ör" (c.f. Helsingör, Lappörarna, Bockören, Kungsör, Skanör) . Linde probably has a dative suffix. Sörberge and Norrberge might well be a dative as well (but might also be a reduced "-berget"), in the vicinity of those you also find some -ede (which might be dative of -ed), and Svedje (which might be a dative of "sved", but not of "svedja"; who knows.) Näse looks very much like a probably dative of "näs" (rather than a nonsense form of "näsa"). However, this too could be a reduced definite form.

6. Adjectives and adverbs that are really datives or -o obliques

ånyo - anew
förvisso - for certain, ("for wit")
åsido - aside
redo - ready - might be a case form

lagom - suitable, to a proportional/suitable extent, quantity or quality'. Definitely a dative form of lag ('law').

stundom - sometimes. Plural dative of 'stund', i.e. 'a while' (c.f. German Stunde, "hour").

bråttom - requiring haste, from bråd ('haste, sudden').

fordom - in ye olden days

enkom - cognate to 'änka' (widow), apparently from an adjective derived from 'one'. It describes something that is 'as if made specifically for' some person, purpose or thing.

länge comes from the accusative form of long ('lång'), and signifies 'for a long time'. This seems to have already been turning into a lexeme of its own in Old Norse.

7. A single example maintaining the previous double genitive -s- and other non-prepositional genitives

In older Swedish - much like cognate icelandic forms today, the genitival -s- went both on the stem and after the definite suffix for definite nouns of the appropriate declension.

Today, this is seen exclusively in the phrase dagsens sanning, the truth of the day.

Some "naked s-genitives" can be found in adverbial roles:
sams: (from the cognate of 'same'), signifies 'in agreement'.
ens - one's - 'even' (mainly in negative clauses or questions, in Finland more widespread than that).
dags ('(it is) time'), from the cognate of 'day'.
strax? ('Soon').
nyss? (new's, signifying 'a moment ago').


8. In compounds

This is one type of position where you actually find a near-productive survival of the case system. Swedish compounds are somewhat tricky with regards to morphology - sometimes, there are linking morphemes (-s-, -e-), sometimes, the stem is used, sometimes the nominative, and sometimes something else. -s- and -e- do look a bit suspicious given what we know of the modern as well as historical genitive (-s-), and -e having been both a masculine nominative marker and a dative marker. I am inclined to think they originate with the historical cases, but their distribution has over time come to reflect the nominal conjugations ever decreasingly.

Some of these are place names: Stugu-näs (cottage.obl-isthmus). Others are verbal compounds, some nominal compounds, some adjectival:

vattumannen - waterman (aquarius)
vattustark - water strong (e.g. too dilute) (dial.)
vattupass - water level (dial. or arch., nowadays often vattenpass),
vattukanna (dial.) - water pitcher

Water is normally vatten in Swedish, and -u is clearly the oblique case form.

skadeanmälan ('skada' , the modern nominative, is a survival of the historical oblique, whereas skade is a survival of the old nominative) - damage report
högskoleprov (skola is similar to skada) - literally "high school (entrance) test", but högskola signifies polytechnics or universities.

sagotant - fairytale-aunt, i.e. some aunt that reads fairytales for children
sagofigur - fairytale-character
sagovärld - fairytale-world
sagotimme - fairytale-hour (the time for reading fairytales at kindergarten)
sagobok - fairytale-book
sagoväsen - fairytale-being, ...

människobarn - human-child,
människoformen - the human shape
människosläktet - the human species

sidospår - side trace, side track
sidojobb - side job
sidoväg - byway
sidoskåp - side cupboard

vargavinter - wolf winter, i.e. a very cold winter
ulvakläder - wolf clothes, in the inversion "får i ulvakläder" (sheep in wolves clothing), that sometimes is used to express someone projecting an exaggerated threat)

mannaminne - "man-memory", a time unit roughly corresponding to the time the oldest generation has been alive

syndaflod - sin flood, the deluge
syndaskrollon - the sin fold (harmonica).

barnatro - child's faith, i.e. the naive and uncomplicated acceptance of religious claims by a kid,
barnavård - child care

 (but barnkläder, barnskötare, barnläkare)


kvinnorättsfrågan (women's rights question)
kvinnokläder (women's clothes)
kvinnosakskämpe (women's issues fighter),

jordavarelse - earth being, earthly creature
jordalivet - earthly life 

nattamat - a small meal during night time
 

(but jordklotet - earth sphere, jordägare, "earth owner")

varuhus, varubod - warehouse, warehut?

salutorg, saluhall, saluföra - market square, market hall, to sell

nämndeman (from 'nämnd', dative?) - committee member

backstugusittare - "hill cottage sitter", a social class in Sweden and Finland in the 18th and 19th centuries. (Don't believe wikipedia when it claims it was a "social class in Sweden"! Swedish wikipedia editors are notoriously unaware of Finland.)

legosoldat - mercenary (from lega, nowadays leja - to employ)

hövdingasäte - seat of a chieftain

nådaår - year of mercy

Dialectal and archaic vocabulary can be added ad nauseum: väga-kant, väga-arbejtar, väga-lag, vattu-drag, vattu-rör, vattu-täkt, vattu-svensk, +tunnebindare*,

* This is weird, because -e there might just be a linking morpheme.

9. Dialectal retentions

Some northern dialects (especially Dalarna, Älvdalen, and occasional places north of there) may have preserved three cases - and intriguingly, at least Älvdalska has lost the genitive, doing the exact opposite of the majority of mainland Scandinavian.

My own dialect is claimed by some speakers to have had a feminine dative pronoun ('hennar') in use as an actual dative until rather recently. Nowadays, we're going in for the Norrlandic loss of case distinction in pronouns, though. Some dialects may well keep a henne/hennar distinction, and I would not be surprised if some dialects maintain "honom" as the dative pronoun, using "han" for nominative and accusative.

10. "Accusativism"

There are several examples in Swedish where the oblique forms have replaced the nominative - skola (from old Swedish skoli, obl. skola), skada (skadi, skada), låga (logi, loga) are three examples. Since -a is identical to the feminine marker, these are hard to spot unless one is aware of them. -e in compounds can, however, be an indicator: högskoleutbildad, högskoleansvarig, skadeanmälan, skadeförsäkring, (however, misleading examples can be found, such as stugetak) .... Ape/monkey - apa - seems to have gone through the same shift even earlier, but we still get apesläktet, the ape family, in some older works.

The loss of Old Norse -R for many nouns may also be a result of a more widespread accusativism (c.f. nom: maðr acc: man, modern Swe: man), ved from {veðR, veð, veði, veðdar}. Whether the F-Swedish dialectal vederlider (for F-Swedish vedlider, for standard Swedish vedbod) originates with nominative veðr or genitive veðar is hard to decide.

Eastern Swedish dialects (i.e. in Finland) have largely undergone accusativism, retaining the oblique case for most nouns at the expense of the nominative, giving a striking difference that is hard to explain as the result of sound changes. The word to the right is the dialectal version expected in large areas in Finland (with possible further vowel changes in the first syllable)

påse - påsa
stake - staka
backe - backa

kaka - kako
ruta - ruto
låga - lågo

In some parts of Ostrobothnia, the adjective as well keeps the masculine accusative morpheme - "en godan kako". (My dialect has replaced the comparative marker -are with the accusative marker instead, so in my dialect, that would signify a 'tastier cake', not a tasty one. Apparently, speakers near the geographical boundary between -an-as-masculine-marker and -an-as-comparative manage to keep trace of what variety you speak.) In southern Finland, the masculine nominative -er on adjectives survives and has often spread quite a bit, being used for feminines as well for some speakers.

 

11. Adjectivified genitives

enahanda, mångahanda, allehanda, varjehanda, tveggehanda, tveggedel, tveggeskafta, treggehanda(!), laga, ...

'Tvegge' and 'tregge' themselves are genitive forms, so 'tveggedel', despite not ending in -a, has a genitive component.

All the 'handa'-adjectives signify the amount of variety something exhibits: "enahanda" means uniform, boring, monotonous. Mångahanda, allehanda, varjehanda - of many/all/every varieties/variety.  Tveggehanda, treggehanda - of two/three types. Tveggedel - two-parted. Tveggeskafta - two-handled/with two shafts.

(d)jävla, (d)jäkla, jädra - genitives of curses, i.e. 'of the devil' with a variety of bowdlerization.


12. Finnish loans from Swedish

It seems Finnish has historically preferred borrowing the oblique form (and basing the Finnish nominative off of it). Thus, many of the Finnish loans do the same a# -> u#-shift. Whether this is due to the dialects doing it first or due to the obliques being more frequent for many words is hard to tell, but given the timespan over which this seems to have been going on, it makes sense to think it's also been a Finnish phenomenon for a longer time than our accusativism has been. Maybe Finnish affected the Finland-Swedes into this accusativism? We do also find exceptions, such as Swedish påse/Old Swedish posi ≃> modern Finnish pussi.

Appendix: Old Swedish Case Morphology in Tabular form

Dove, a feminine noun, here with some of the alternative suffixes omitted (-om/-um, -umin/-omen, etc seem to have been in free variation). Modern Swedish genitives and nominatives are also included for comparison:
dūva (dove), f:

sing indefsing defplur indefplur def
nomdūva (duva)
dūvan (duvan)
dūvur (duvor)
dūvuna(r) (duvorna)
accdūvudūvuna
dūvur
dūvuna(r)
datdūvudūvunni
dūvum
dūvumin
gendūvu (duvas)
dūvunna(r) (duvans)
dūva (duvors)
dūvanna (duvornas)
  
 
As we can see, acc/dat/gen are conflated in the singular in the indefinite singular. There are, however, some other conflations, each marked by colour coding. Notice that 
posi (sack) conflates acc and gen in the plural wih acc/dat/gen in the singular.
posi (sack), masc:

sing indefsing defplur indefplur def
nomposi (påse)
posin (påsen)
posar
posarnir
accposa
posan
posa
posana
datposa
posanom
posom
posonom
genposa
posans
posa
posanna  
 
For some nouns, nom-acc-dat are conflated in the indef. sg., and nom-acc are conflated everywhere. Some other conflations may occur, e.g. acc/dat-conflation or dat/gen-conflation. Neuters conflate nom and acc, sometimes nom-acc-gen.

The -r- in modern 'duvorna' (and other similar nouns) is probably an orthographical innovation to make the forms more consistent. It is pronounced in but a few dialects.

Adjectives

There are 48 slots in the adjective paradigm. Luckily, there's a lot of syncretism, but ... first: the strong and weak adjectives. Weak adjectives tend to be used with definite nouns: the strong man, the strong one.  Here, I used "strong" to emphasize that there is no semantic significance to the names strong and weak: they could just as well be called red and blue adjectives or medieval and oily adjectives. Strong adjectives are used for indefinite nouns, and for complements. Exactly when they started operating as complements - or if it's west Germanic that has lost congruence on complements - is unclear to me.

The strong adjectives have this pattern or something very much like it:


masc sg        
fem sg     neut sg            
masc pl        
fem pl        
neut pl




nom -er, -
-  
-t
-ir, -er
-ar
-  




acc -an
-a
-t
-a
-ar
-  



dat -um, -om
-(r)i
-u, -o
-um, -om
-um, -om
-um, -om




gen -s
-(r)ar
-s
-(r)a
-(r)a -(r)a




In modern Swedish, this has simplified to nothing for reale sg, -t for neuter, -a for plurals.

The weak ones are like this:


masc sg        
fem sg     neut sg            
masc pl        
fem pl        
neut pl




nom -i/-e     -a      
-a
-u, -o -u, -o -u, -o



acc -a
-u, -o
-a
-u, -o -u, -o -u, -o



dat -a
-u, -o -a
-u, -o -u, -o -u, -o



gen -a
-u, -o -a
-u, -o -u, -o -u, -o



In modern Swedish, these have become -a. For all forms. Except masculine nouns can take -e, and there are some other complications in how -e is used. In some dialects, the -e forms also appear with (masculine) plurals, and in some dialects, the -u forms have won out instead.


Appendix II: additional examples to incorporate

One major issue with writing this is that new examples keep popping up, both through introspection and through reading. Some of these are examples I've not even been aware of, others are ones that I am reminded of when encountering them, and some of them are ones that just pop up out of introspection, and when looking them up they turn out to exist. A few false starts of that kind have happened as well, but are filtered out by actually looking them up. This might indicate that this is "quasi-productive" in Swedish, though?

till synes

nåd, nåde, nåder

i rättan tid

gå i sömne

urbota

båda/bägge as case distinction seems to survive in some regions

i sin prydno

huruvida/såtillvida

i blindo

ge vika ?

bärsärkagång

mörksens gärningar

himlavalvet and "himla tider"

inalles (german loan?)

så här års

 förstone/sistone/åtminstone/


1 comment: