Norwegian: distribution of seg, seg selv, etc.

[tetsugakusha43]

First off, I would like to know if each of the sentences in (1-2) sound acceptable to you. In case they do, does the focused phrase in each case (ham, ham selv, seg, seg selv, sin) refer to (i) just Jon, (ii) just Karl, (iii) either Jon or Karl, (iv) a third dude, or (iv) all of the above?

(The examples may be very awkward with the verbs I chose for unrelated reasons; I am open to suggestions/corrections, obviously!).

Topicalization
(1a) Om ham sa Jon at Karl aldri ville slutte å snakke
(1b) Om ham selv sa Jon at Karl aldri ville slutte å snakke
(1c) Om seg sa Jon at Karl aldri ville slutte å snakke
(1d) Om seg selv sa Jon at Karl aldri ville slutte å snakke
(1e) Om moren sin sa Jon at Karl aldri ville slutte å snakke

Clefts
(2a) Det var ham Jon sa at Karl beundret
(2b) Det var ham selv Jon sa at Karl beundret
(2c) Det var seg Jon sa at Karl beundret
(2d) Det var seg selv Jon sa at Karl beundret
(2e) Det var moren sin Jon sa at Karl beundret

 

[raumar]

Hello, Tetsugakusha,

Before we start to answer your questions, I think it would be good if you can clarify what the sentences under (1) are supposed to mean. At first glance, most of them don't make any sense. Take (1e), for example. The natural understanding of this sentence in Norwegian would be that it means "Jon sa om moren sin at Karl aldri ville slutte å snakke". But that does not make any sense. If Jon says "Karl wouldn't stop talking", Jon is obviously talking about Karl, not about anybody's mother.

Then I realized that you maybe meant to say "Jon sa at Karl aldri ville slutte å snakke om moren sin". This would make sense, but if that is your intended meaning, I am afraid that your original word order is Yoda-speak.

 

[myšlenka]

This reeks of generative syntax, binding relations and island constraints. Intuitions for grammaticality judgments for these cases are probably very subjective. Personally, I have no issue with extracting a simple DP verb complement and moving it into the topic position of the matrix clause. However, when the verb complement is an entire PP (like all your sentences in 1), things get more messy. I am not sure I like pied-piping the P. It seems to me that things improve if the P is stranded in the embedded clause. This is probably because topicalising the entire PP almost creates a garden path sentence where my brain wants to interpret the PP as belonging to the matrix verb instead of the embedded verb. Thus, a processing issue that gets even more complicated with anaphoric DPs. I am not sure if a stranded P in the embedded clause is outside the scope of what you are investigating, but it improves things massively. I took the liberty to change the verbs a bit to illustrate the effect (topicalising pronominal DPs like this requires focus, hence the capital letters):

a) HAN tror John at søstera stoler på.
b) *På HAN tror John at søstera stoler.

Also, since you are looking at binding of pronouns and anaphoric expressions, you should choose the embedded verbs more carefully (or perhaps that is what you have done already). In the class of verbs that are reflexive by default, some of them can have the reflexive semantics reinforced with selv (han vasket seg selv), while other reflexive verbs with opaque semantics are either incompatible with selv (*han gruet seg selv) or lose semantic opacity with selv (han liker seg versus han liker seg selv, i.e. the addition of selv imposes transparent semantics). As for transitive verbs, they are generally incompatible with reflexive seg unless you also add selv.

c) *Han beundret seg.
d) Han beundret seg selv.

Thus, the phrase structure in both 1c) and 2c) from your set of sentences would probably be judged as ungrammatical (even with P-stranding) due to the argument structure of the embedded verb.

That being said, I can try to answer your question.
1a+2a) A third dude
1b+2b) Jon
1c+2c) Ungrammatical
1d+2d) Karl
1e+22) Karl

Or something like that