The reduction of denn to n and some of its consequences Josef Bayer Universität Konstanz I Introduction The German particle denn is interesting for various reasons It depends on interrogative force ID: 830477
Download The PPT/PDF document "SaRDiS 2020, 20-21.11.2020" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
SaRDiS 2020, 20-21.11.2020The reduction of denn to -n and some of its consequences
Josef
Bayer
Universität Konstanz
Slide2I. IntroductionThe German particle denn is interesting for various reasons. It depends on interrogative force;
It can under special conditions appear in dependent non-interrogative clauses;
It gets reduced in spoken language and dialects to the weak forms ‘
dn or ‘n. The present talk will concentrate on the last property.
The reduction of denn to -n and some of its consequences
2
21 November 2020
Slide3II. Distribution and semantic contributionDenn is optional in questions. (1) a. Wer hat eingekauft? / Ist er im Urlaub? b. Wer hat denn eingekauft? / Ist er denn im Urlaub?
Denn
requires interrogative force; it is out in declaratives or imperatives. (2) a. *Klaus hat denn eingekauft. b. *Kauf denn ein! [Watch out: Some northern speakers allow denn
in the sense of “dann”, “danach
”.]
The reduction of
denn
to -n and some of its consequences
3
21 November 2020
Slide4What is its semantic contribution?Denn is "expressive" rather than "descriptive/truth-conditional" (3) denn
(
) is felicitous if (i) is a question, and (ii) the expected true answer p updates the common ground CG that the speaker S shares with the addressee in the actual context c in such a way that S is in some way concerned about p with respect to c. The context c can be provided by an antecedent sentence: A: Klaus will Urlaub in Sizilien machen. B:
Hat er
denn genügend
Geld?
But also by a non-linguistic situation, e.g. the children have flooded the living room in order to have their toy boats float on the water.
Daddy detects this and asks:
Seid ihr denn wahnsinnig geworden?
Here,
denn
is quasi anaphoric to the situation, the mess whose existence is in the CG of the interlocutors.
The reduction of denn to -n and some of its consequences
4
21 November 2020
Slide5Excludes out-of the blue questions (in Konstanz: Wo ist denn hier der Bahnhof? vs. #Wo ist denn in Bordeaux der Bahnhof?)Excludes emotionless questions (new acquaintance:
Wo
wohnen Sie denn? vs. immigration officer: #Wo wohnen Sie denn?)
leaves emotional dimensions open (Wie geht es dir denn jetzt?
Was fällt Dir denn ein?)
not necessarily confined to information-seeking questions (
Was
schaust
du
denn
so
dumm
!
Wo wird
er das Geld denn schon her haben!) Paraphrase: denn “under the actual circumstances”.
The reduction of denn to -n and some of its consequences
5
21 November 2020
Slide6III. SyntaxLike other DiPs, denn has a fixed position. It is in the highest middle field, preceding other DiPs
as well as higher adverbs. Elements to its left, if any, have topic status (
aboutness
as well as discourse topics are ok). Denn precedes vP, and also NegP(4) Wer ist (denn) bei dem Angriff (denn) nicht (*denn) verletzt worden?Word order variations are the result of scrambling to its left. The DiP
itself never moves. This distinguishes DiPs from their adverbial siblings; s. dann
, which can be moved.
The reduction of
denn
to -n and some of its consequences
6
21 November 2020
Slide7The assumption is that denn, like other DiPs, is a functional head along with other functional heads like v, T, C, Neg etc. If so, the particle (
Prt
°) heads a particle phrase (
PrtP). topic(s)1 [PrtP Prt° [AdvP Adv* [VP ... t1 ... ]]]
Evidence
for head status: denn reduces to the
clitic
element ‘
n
. Such reduction affects heads and not XPs.
Compare the adverb
vielleicht as in Vielleicht ist der Klaus im Urlaub
or Der Klaus ist im Urlaub vielleicht with the DiP
vielleicht, particularly its reduced form
leicht
in Austrian/Viennese: Bist leicht deppert. Cf. *Leicht bist deppert, *Bist deppert leicht.
[There is independent evidence for head status which cannot be presented here].
The reduction of denn to -n and some of its consequences
7
21 November 2020
Slide8PronounsPronouns scramble to the left of the DiP obligatorily, unless they bear contrastive stress. (5)
a. Hat er denn was gesagt?
has he DENN anything said “Did he say anything?” b. *Hat denn er was gesagt? c. Hat denn ER was gesagt? “Was it HIM (in contrast to HER) who said anything?”Pronouns are prototypical topics. It goes without saying that
unstressable elements (es
, man) and clitic
forms cannot stay in the scope of the
DiP
either.
(6) a. Hat es denn geschneit?
has it DENN snowed
“Did it snow?”
b. *Hat denn es geschneit?
(7) a. Kann man denn hier rauchen?
can one DENN here smoke “Is one allowed to smoke here?” b. * Kann denn man hier rauchen? The reduction of denn to -n and some of its consequences821 November 2020
Slide9It is so far unclear why DPs with topic status may stay below denn but not pronouns. (8) a. *Wollte denn der Manfred ihn mitbringen?
wanted DENN the M. him
along.carry
b. Wollte ihn denn der Manfred mitbringen? Although much more could be said, this should conclude the sketch of the grammar of denn in Standard German. We turn now to further developments in the grammar of denn that have taken place in the Bavarian dialects and maybe other dialects as well.
The reduction of denn to -n and some of its consequences
9
21 November 2020
Slide10IV. Origin and developmentOrigin: from Idg. root *to which had a deictic meaning. First locative then also temporal adverb
danne
/
thanne ("from then", "from there"). (cf. the relation to English then). Then also causal denn ("because") – Klaus ist blass, denn er hat die ganze Nacht
gefeiert ("Klaus is pale because he has celebrated the whole night"). OHG anaphoric usage of thanne
. ("under these circumstances", "thus").
(9) a.
Ther púzz ist filu díofer, war nimist thu thánne ubar tház
the well is much deeper where take
you
then
over the
wazar flíazzantaaz [Otfried II, 14, 29f.]
water running “The well is very deep, so where will you then take running water?”The reduction of denn to -n and some of its consequences1021 November 2020
Slide11b. Thiu óugun sie imo
búntun
[...]
joh frágetun ginúagi, the eyes they him bound and asked forcefully wér inan thanne
slúagi [Otfried
IV, 19, 73f.] who him then beat
“They covered his eyes and asked him who then beat him”
Denn
in modern German retains this
anaphoricity
.
The reduction of
denn to -n and some of its consequences
11
21 November 2020
Slide12(10) Grammaticalization path (Abraham, 1991; Wegener, 2001) LOCALISTIC > TEMPORAL > LOGICAL > ILLOCUTIVE / DISCOURSE-
FUNCTIONAL
The reduction of
denn to -n and some of its consequences1221 November 2020
Slide13The fate of denn in Bavarian: Bavarian is a spoken not a written language. A most prominent property is that Bavarian has, unlike Standard German, genuine clitics (comparable to those of Romance). Phonological integration yields processes unknown in the standard language.
(11) a.
SG: Wenn ich es dir doch sage if I it you DOCH say “if I tell you“ b. BAV: Wenn-a-s-da doch so:g
(12) a. SG: wie ich mich umdrehe
as
I me
around.turn
“
as I turn around”
b. BAV: wia
-r-a-me umdrah
[consonantal epenthesis]
The reduction of
denn to -n and some of its consequences1321 November 2020
Slide14Cliticization has grabbed also denn: denn > ‘n
(13)
Wo
wohnst-n du? where live -N you “Where do you live?” (14) Wann hod -a -s-da -n zoagt? when has-he-it-you-N shown
“When did he show it to you?“
The reduction of denn to -n and some of its consequences
14
21 November 2020
Slide15–n follows the clitic pronouns. Permutations are strictly out. –n cannot cliticize to an XP [no phrasal
clitic
!]
(15) a. *Wann hod-a-s-n-da zoagt? b. *Wann hod
-a-n-s-da
zoagt?
c. *Wann is d’Sophie -
n
in des Haus neiganga?
when is the-Sophie-N in this house
in.gone
“When did Sophie enter this house?”
The reduction of denn to -n and some of its consequences
15
21 November 2020
Slide16The full form denn is not very popular in uncorrupted varieties, and tends to be replaced by various forms deriving from the adverb nachher :nacher, no,
noucha
,
nou etc. (16) Wann is-n d’Sophie nacher in des Haus neiganga? when is-N the-Sophie NACHHER in this house in.gone
“When did Sophie enter this house after all?”
Notice that –n
and
nacher
can
cooccur
, and
nacher
is an XP; as such it can follow another XP.
The reduction of denn to -n and some of its consequences
16
21 November 2020
Slide17The landing site of –n is very clearly the Wackernagel position, usually called the C-position, which may be enriched by clitic pronouns if there are any. (17) [
CP
wh
1 [C’ [C° Vfin] [(CL*) –n … [VP ... t1 ... ]]]]
[C+CL*+n
] form an X° complex
The reduction of
denn
to -n and some of its consequences
17
21 November 2020
Slide18The reduction of denn to -n and some of its consequences1821 November 2020If this was all there is, the syntax of –
n
could be described as an allophonic variant of the syntax of
denn in Standard German. However, Bavarian -n is at variance with the full form denn in at least two respects: (A) –n is obligatory in wh-questions, unlike
denn.(B) –
n is deprived of its semantics
in
wh
-
questions, unlike
denn
.
(18) a. Wos
host *(-n)
g’sagt? Weiß
(2002)
what have-you –N said “What did you say?” b. Wos faid *(-n) dene ei
? what falls –N these in “
What comes to their mind?”
c. Wos schaust *(-
n
) so g’spassig?
what look-you -N so funny
“Why do you look so strange?”
Slide19Aside from phonological weakening, both obligatorization (A) and semantic deprivation (B) are typical signs of grammaticalization; s. Lehmann (1995). (A) and (B) are obviously related: While in Standard German
denn
is optional and as such makes a semantic contribution, -
n is obligatory in all sorts of contexts and therefore applies to wh-questions across the board. In other words, it makes no semantic contribution. Givón (1976): "Grammaticalization rips off the lexical features until only the grammatical features are left“Consequences:
(i) All Bavarian wh
-questions, including out-of the blue
wh
-
questions, show –
n
.
(ii) Bavarian out-of the blue
wh-questions do not show
nachher, nachher
being a meaningful optional element; [same as denn
].
(iii) –n is the morphological exponent of a purely formal feature [“denn minus its semantics”]The reduction of denn to -n and some of its consequences1921 November 2020
Slide20Recall that –n appears obligatorily only in wh-questions. Since –n is part of the featural
X° amalgam [C+(CL*)+
n
] in C°, C° inherits the relevant wh-feature. As a consequence, C° agrees overtly with the wh-operator in SpecCP. For Bavarian, the grammaticalization path in (10)
appears to be extended as in (19)
(19)
Grammaticalization
path of
denn
for Bavarian
LOCALISTIC > TEMPORAL > LOGICAL > ILLOCUTIVE / DISCOURSE FUNCTIONAL > WH-
AGREEMENT MARKERThe reduction of denn
to -n and some of its consequences
20
21 November 2020
Slide21Are reassurance questions (echo questions) counter-examples? (20) A: I hoass
Notburga
“I am called Notburga” B: WIA hoasst du? / *WIA hoasst-n du?? “You are called WHAT??” B‘: Du hoasst WIA?? / *Du hoasst-n WIA
?? “You are called WHAT??”
In an information-seeking wh
-
question
wh
1
[
… t1 …], the speaker holds the presupposition
’ in which the variable is existentially closed: wh1
x [
’… x …]. In a reassurance wh-question, the speaker has already some sort of value for the variable and expresses his/her dissatisfaction or disagreement with this value. If so, reassurance questions may not even be interrogative in the narrow sense. The reduction of denn to -n and some of its consequences
2121 November 2020
Slide22V. A digression into Rheto-RomanceAccording to Hack (2009; 2014), the particle pa or its reduced form a
, derived from
Latin
post, „behind, after“, has lost its meaning and has become obligatory in wh-questions of the Dolomite dialects of Badiot, Marèo and Gherdëina (Grödnertal) (21)
Badiot
Ci fajés
pa
?
what do.2PL PA
“What are you doing?”
The reduction of
denn to -n and some of its consequences
22
21 November 2020
Slide23(22) Gherdëina Cie fajëis’ a
sën
? what do.2PL=PA now “What are you doing now?” In polar questions of Badiot and Marèo, pa is optional. Interestingly, the same is true for –n in Bavarian polar questions.
The reduction of denn to -n and some of its consequences
23
21 November 2020
Slide24The reduction of denn to -n and some of its consequences2421 November 2020However, Gherdëinan
has developed further than
Badiot and Maréo in that pa has become obligatory in polar questions as well. (23) Gherdëina a. Vën =iel pa
ence Tone?
come.3SG=SCL PA also Tone
“Is Tone coming as well
?”
b. Ne
uniëis
’
a
nia? NEG come.2PL=PA NEG
“Don’t you come?”
Slide25VII. DropWe know from spoken German that in declarative clauses the topic in SpecCP
can
be dropped under certain discourse
conditions, e.g.(24) a. Der steht schon vor der Tür he stands already before the door
Den
kenn ich nicht
him
know I not
The reduction of denn to -n and some of its consequences
25
21 November 2020
Slide26(ii) We know from Bavarian that the 2nd pers. sg. and pl. pronoun and occasionally also the 1st pers. pl. pronoun can be dropped, e.g. (24) b
.
Bi-
st du no:s woan? “Did youSG get wet?”are yousg wet become
Seid-ts es
no:s
woan
?
“
Did
you
PL
get wet?”
dass-ts es
no:s
woan seid-ts “that youPL got wet”that youpl wet become are
The reduction of denn to -n and some of its consequences26
21 November 2020
Slide27VIII. Wh-DropGiven the obligatory presence of –n, the exponent of the u
Q
feature in C, it is expected that a
wh-operator in SpecCP can be dropped as well. This expectation is indeed fulfilled: (25) a. Wos is-n do los? “Was ist hier los?” What’s going on here? b. [ _ ] is-
n do los?
(26) a. Wos
dea
-
ts
-
n
es do? “Was tut
ihr da?” What are you doing here?
b. [ _ ] dea
-
ts-n es do?The reduction of denn to -n and some of its consequences2721 November 2020
Slide28Wh-drop is never possible if –n is missing. (27), (28) cannot be interpreted as questions. (27). *[ _ ] is do los? (28). *[ _ ]
dea-ts
es do? Thus, -n is crucial in recovering the featural content of the wh-operator. The reduction of denn to -n and some of its consequences
2821 November 2020
Slide29Was is the only wh-operator that can be dropped. (29) a. Wer hockt-n
in da Stum drin?
“Wer sitzt im Wohnzimmer drin?”
Who is stitting in the living room? b. *[ _ ] hockt-n in da Stum drin?
Why? C encodes no features beyond uQ
. Thus, semantic restrictions for person, time, place, manner etc. would fail to be recovered by –n. The morphological default form of
wh
is the underspecified pronoun
was
. [s. A.
Jäger
, 2000].
The reduction of denn to -n and some of its consequences
29
21 November 2020
Slide30A caveatWh-drop is not limited to Bavarian. Some languages and dialects without a clitic particle allow it too. (30)
a. Was ist
denn
das _ für’n komischer Vogel? b. [ _ ] ist denn das _ für’n komischer Vogel? c. *[ _ ] ist das _ für’n
komischer Vogel?
Ruhrdeutsch; Andreas Trotzke (p.c.)
(31) a. Wat heb je _ voor een boek gekocht?
what have you for a book bought
“What kind of book did you buy?”
b. Wat heb je
nou
_ voor
een boek gekocht?
c. [ _ ] heb je nou _ voor een boek gekocht? d. *[ _ ] heb je _ voor een boek gekocht?
Dutch; Marcel den Dikken (p.c.)
The reduction of denn to -n and some of its consequences
30
21 November 2020
Slide31Informal testing of 13 non-Bavarian / non-Southern speakers with sentences of type (32) Context: Klaus knows that Konrad is not the best cook; he throws a critical look in Konrad’s pot and asks him:
(32)
a. Hast'n DU da gekocht? b. Hast denn DU da gekocht? c. Hast DU denn da gekocht? d. Hast DU'n da gekocht?This test revealed that two speakers did not accept any of them whereas the remaining 11 speakers gave the best grade to type (32a). Less people accepted (32b) or judged them as worse. Hardly anyone accepted type (32c). (32d) was not an option for anyone.
The reduction of denn to -n and some of its consequences
31
21 November 2020
Slide32This shows rather clearly that the use of –n is close to the one in Bavarian, and that even the full form denn is on the way to be integrated in the C°-complex. Wh-drop is impossible if –n/denn
is not adjacent to the C°-complex. This is in sharp contrast to the rest of the syntax of
denn
by which full DPs with topic status can freely precede denn. The reduction of denn to -n and some of its consequences3221 November 2020
Slide33IX. ConclusionThe Q-sensitive discourse particle denn shows an interesting development that can be described as steps of grammaticalization. These steps involve more than the phonological reduction
(33)
dann
> denn > d’n > –n In Bavarian wh-questions, -n has become obligatory and has lost the residues of its discourse function of making reference to some common ground between speaker and hearer. It seems to be justified to say that it has shifted from a Q-sensitive discourse particle to a wh agreement marker. In its latter function, it enables wh-
drop.The reduction of denn to -n and some of its consequences
33
21 November 2020
Slide34Vielen DankThank You
The reduction of denn to -n and some of its consequences
34
21 November 2020