A dependency grammar represents the structure of a clause as a set of pairs of words, such that the first "depends" on the second. My question is ultimately how loops are prevented in such a representation (a loop would be a case in which, say, A is a dependent of B, B a dependent of C, and C a dependent of A).

Specifically: It seems to me that the analysis of relative clauses could create such a situation: a relative clause is an attribute (hence a dependent) of a noun; it is (let's assume) introduced by a relative pronoun and the verb of the relative clause would have to depend on the relative pronoun, since the pronoun introduces the relative clause and is responsible for the clause typing. Simultaneously, the relative pronoun functions as the subject or object of a relative clause, hence would be a dependent of the verb.

I consulted a dependency grammar of German (Eroms: Syntax der deutschen Sprache. Walter de Gruyter 2000). There, the author proposes to split the relative pronoun into two elements. One of them acts as the head of the relative clause, the other as the dependent of the verb. So the description of the phrase "dogs, which bark" would read (if <1,2> is read as "1 depends on 2"): <wh-, dog>, <bark, wh->, <(wh-)ich, bark>. So "which" would have been split into "wh-" as a kind of relative clause complementiser and "(wh)ich" as a subject pronoun (cf. p. 292 in that book).

This analysis strikes me as "generative grammar in disguise" (because of the introduction of invisible additional elements), and as insufficient. One detail: in German dialects you can get a combination of relative pronoun plus conjunction, like: "Leute, die wo sowas machen". Generative models can handle this in their X-bar-theory as specifier plus C-head. But doesn't the splitting analysis lead one to expect (at best) a structure "wo die"? Eroms alludes to the existence of those dialect data as support for his analysis, but it is hard to see how they actually support it. ...This is a minor point here, but I'm following a request to provide the details. The basic question was:

Has this particular dependency analysis been defended elsewhere, and are there any alternative analyses around in dependency grammar that have other means of avoiding a loop?

  • 1
    analysis of German etc.
    – Lambie
    Commented Oct 1, 2023 at 18:02
  • 1
    The theoretical question is independent of German, only the example and the origin of the analysis was German.
    – Alazon
    Commented Oct 1, 2023 at 19:26
  • 1
    I intend to answer this question. However, I am not sure what is meant by "a combination of relative pronoun plus conjunction". An example of this would help. I suggest editing the question to provide such an example. Commented Oct 2, 2023 at 20:44
  • 1
    Question text was edited accordingly.
    – Alazon
    Commented Oct 2, 2023 at 23:05

1 Answer 1


There are at least four basic analyses of relative clauses that one encounters in the DG literature. These four analyses are illustrated with the next dependency trees/graphs of the noun phrase the people who you have called:

enter image description here

The analysis in (1d) is the split analysis alluded to in the question. It is the one with the longest tradition, going back at least as far as Tesnière (1959); it is also the analysis preferred in Word Grammar, and as pointed out in the question, by Eroms (2000). The analyses in (1a) or (1b) are the ones that one encounters in the Universal Dependencies (UD) or Surface- Syntactic Universal Dependencies (SUD) frameworks, and in this regard they are the ones that enjoy the greatest prominence at present.

The analysis in (1c) is the one I prefer and have argued for in my writings (e.g. Osborne 2019). I think there are solid linguistic arguments revealing that the relative pronoun should, when it is present, be viewed as the root of the relative clause; it functions like a subordinator – there are important details of my account that I am omitting from tree (1c), though; simplification makes the competing analyses more comparable.

I agree with the implication in the question, namely that (1d) contains a cycle, a “loop”, and is therefore problematic. Many DGs reject cycles and would therefore reject (1d) precisely due to the cycle it contains given that the relative pronoun appears in two distinct hierarchical positions. I also agree with the critique that the split analysis is problematic because it relies on a null element. Traditionally, DGs eschew the null elements of the sort frequently encountered in Chomskyan grammars. They argue that the existence of such null elements is non-falsifiable and hence scientifically dubious.

Concerning the example Leute, die wo sowas machen ‘people who do such a thing’. I am unfamiliar with such cases. My inclination would be to view die wo as a complex element, perhaps occupying a single node in the tree. In this regard, it might be better to analyze it as a one word diewo, or as a word and a clitic, i.e. die-wo. In any case, it would constitute a compelling argument for the split analysis only in the event that something can intervene between die and wo, which I am guessing is not possible. I also think that the DG analysis of such cases is no better or worse off then an X-bar approach. There is no harm done if one chooses to split the subordinator relative pronoun into two nodes, each being occupied by a separate element that is phonologically present. Doing so would be akin to two-part prepositions, which occur frequently, e.g. out of, in to, on to, etc.

To summarize, dependency syntax is not monolithic. There are competing analyses and one chooses the analysis that one thinks is best. This is common practice in the study of syntax in general, of course. One thing that I think many DG people would dislike, though, is being accused of producing a syntactic analysis that is “generative” in the Chomskyan sense associated with phrase structures.

Addendum The comments to my original answer are motivating this addendum. I mentioned that (1c) omits important details. I now include those details. Tree (1c) becomes (1c’): enter image description here

The dashed dependency edge indicates that who is not the governor of have, and the dangling dependency edge identifies called as that governor. There is hence a long-distance dependency of a sort present, extending from called up to who. I pursue this sort of approach for all long-distance dependencies. To provide another example, here is my analysis of an instance of wh-fronting: enter image description here There is a consistency to the use of the dashed and dangling dependency edges. The dashed edge indicates that the head word is not the governor of the dependent, and the dangling edge identifies which word is that governor.

The approach to dependency syntax just illustrated with these two trees is unlike many other DGs. It is an entirely projective DG, that is, projectivity violations never actually occur. One might point out that the approach is akin to movement-based accounts. I prefer to conceive of these long-distance dependencies in terms of feature passing (think slash mechanism of GPSG), though, which means I advocate for a monostratal approach to syntax –- there are no movements or transformations, but rather just surface structures.

My DG is sometimes characterized as dependency-based GB, that is, in terms of dependencies rather than constituencies. I disagree with this characterization, though. I prefer to characterize it more as a dependency-based version of HPSG.

Finally, the critique of DG in the comments is fair and unfair, in my view. It is fair in that many DGs do ignore critical information of the sort that I include in terms of the dashed and dangling dependency edges. It is unfair in that dependency syntax is much simpler than phrase structure syntax. Dependency syntax can produce plausible analyses of sentence structures with much less effort than phrase structure syntax and in this regard, it is preferable.

  • 1
    (a) Thank you for working out such a detailed answer! By and large, the four solutions represent two groups. 1) The relative pronoun is a subordinator, then you have trouble representing the gap in the verb's valency, or 2) the relative pronoun is a complement of the verb, then you have trouble representing the clause type. Hence, relative clauses seem to show some weaknesses of DG. The problem confirms my overall impression of DG that it tends simply to omit any information that it finds difficult to represent....
    – Alazon
    Commented Oct 3, 2023 at 10:30
  • (b) In the latter solution, you would have to stipulate that when a verbal node depends on a noun, this triggers an analysis as a relative clause. The whole thing reflects the problems with taking the predicate as the root of the clause; notably Eroms introduces an abstract clause-type feature as the root of the clause instead. It seems you would need the same move for subordinate clauses that lack a conjunction, like most relative clauses. But then, there are also relative clauses with conjunction (perhaps English "that"), and do they have a totally different analysis then?
    – Alazon
    Commented Oct 3, 2023 at 10:33
  • (c) The last structure is interesting in that it connects one word to two syntactic nodes simultaneously. This is exactly how you describe movement in Transformational Grammar...
    – Alazon
    Commented Oct 3, 2023 at 10:33
  • Postscript to (b): The second solution is interesting in that it takes the finite verb (any finite verb) as the head of the relative clause and the rel-pro. as a special type of dependent. In generative terms, this means the rel-pro is a specifier of C, with C being the head of the rel-clause. But unlike the CP-analysis, you have no distinction between the V and the C. I conclude that the generative analysis fares visibly better than the standard DG solutions. Of course, the DG structures could be enriched accordingly...
    – Alazon
    Commented Oct 3, 2023 at 10:40
  • 1
    @Alazon, I will add an addendum to my answer in order to address some of your comments. I will not have time to do this for a couple of days, though. Commented Oct 5, 2023 at 0:48

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.