Quirky Subjects
Non-nominative subjects in Icelandic
by C Ryan Moniz
original research· summer 2018
updated & published· harvest 2021
Some Germanic languages, such as modern Icelandic and Old English, exhibit a behavior referred to ask “quirky subjects,” wherein the subject of a sentence takes a case other than the nominative case — the case expected of a subject in these languages. Here, I demonstrate the identification of quirky subjects in both modern and Old Icelandic.
What is a subject?
In order to identify quirky subjects, we must first understand how to identify a subject. A subject is a grammatical relation which:
- acts as the primary actor of typically transitive verbs
- acts as the single argument of intransitive verbs
- is marked by a particular case or lacks such marking in contrast to marked cases
- is conditioned by agreement with the verb (e.g. we see_ versus she sees)
- is capable of being deleted in certain positions, such as coordinate structures (e.g. they baked bread and
theyate it) - is capable of coreference with reflexive pronouns
- in languages with question, negative, or topical inversion, subjects invert adjacent to the verb
Depending on the language in question, a subject will meet a variety, though not necessarily all, of these criteria.
What is a “quirky” subject?
“Quirky subjects,” also known as “non-nominative subjects” or “oblique subject-like arguments,” are arguments for certain verbs which take a case other than the nominative. They may pass some of the criteria for syntactic subjects given above, while potentially failing others. They are still a debated category for many languages — the clearest case for their existence is modern Icelandic.
A confounding factor
Once the criteria for identifying a subject is established, it may seem straightforward from that point on to identify a subject and then check whether it is marked (or unmarked) for a nominative or non-nominative case in order to determine its statnus as “quirky” or not. However, there is a confounding factor that makes identifying quirky subjects challenging: the variance of impersonal verbs. Very often, verbs which take quirky subjects are listed as “impersonal verbs.” In many languages, such as Spanish, certain psychological/emotional verbs appear in a construction with the following elements:
- Y (in non-nominative case)
- V (verb agreeing in person & number with X)
- X (in nominative case, agreeing in person & number with V)
For example, consider the following Spanish sentence:
- meY gustaV la florX • “I like the flower” [literally: “The flower is pleasing to me”]
As indicated by the labels and literal translation, such sentences do have a true morphological and semantic subject in the nominative case which has simply been moved to a position after the verb. The verb gustar “to be pleasing” agrees in person and number with the subject X la flor.
This construction, which also appears in many Germanic languages, is a potential problem in identifying quirky subjecs because many clauses containing quirky subjects, especially dative subjects, do morphologically match this impersonal verb structure, with the key difference being that the non-nominative argument (analogous to me in the Spanish example) behaves syntactically like a subject. In order to demonstrate that a quirky subject is indeed a subject, there are a few syntactic tests which can be applied.
Subject tests in Modern Icelandic
In order to demonstrate the process of determining a quirky subject, I will first present some of the key subject tests with modern Icelandic verbs which take quirky subjects.
Reduction in coodinate structures
One test which can be applied to determine the subjecthood of an argument suspected to be a quirky subject is to see whether it is grammatical to speakers of that language to delete it when it is part of a coordinate structure with another (usually nominative) subject. Consider the following examples with the phrasal verb verða óglatt “to become queasy,” which puts the argument who is becoming queasy in the dative case:
- Honum varð óglatt. • “HeDAT became queasy.”
- → Hann kom og
honumvarð óglatt. • “HeNOM came andheDATbecame queasy.”
From this test, we can see that the dative honum is being treated as syntactically equivalent hann in the nominative. We can compare the results of this test to an analogous impersonal verb in modern German:
- Ihr wurde übel. • “HeDAT became queasy.”
- → Er kam und ihr wurde übel. • “HeNOM came and heDAT became queasy.”
In contrast to the situation in Icelandic, the dative ihr in German is grammatically required, i.e. it cannot be deleted from the impersonal werden übel, indicating that it does not have the same degree of ”subjectness” as the Icelandic honum.
Raising to object
Object raising is a phenomenon where the subject of an embedded conjugated verb is moved to be the object of another verb, and the former embedded verb is put into the infinite. To give an example, look at these sentences for a more typical verb (one which does not take a quirky subject) in English and in Icelandic:
- English
- I believe that sheNOM is strong.
- → I believe herACC to be strong.
- Icelandic
- Ég tel að húnNOM sé sterk.
- → Ég tel hanaACC vera sterka.
In both English and Icelandic, the original nominative subject of “be strong” becomes an accusative object of the verb “believe.” The object raising test checks whether non-nominative arguments can similarly be raised to another verb. Below is an example of Icelandic exhibiting this behavior, with the verb leiðast “to be bored by something,” which takes a dative subject:
- Ég tel að henni leiðist bókin • “I believe that sheDAT is bored by the book.”
- → Ég tel henni leiðast bókin. • “I believe herDAT to be bored by the book.”
Interestingly, the dative subject does not change to an accusative object, as happens with more typical subject-to-object raising, but remains in the dative.
Subject-auxiliary inversion
There are many contexts in Germanic languages wherein the subject of a verb will switch places with an auxiliary (helping) verb. In English, this happens with, for example, questions (e.g. “You are hungry.” → “Are you hungry?”); however, in Icelandic and many other Germanic languages, one trigger for subject-auxiliary inversion is the topicalization of time, i.e. when a time phrase is moved to the beginning of the sentence. Icelandic non-nominative subjects have demonstrated this behavior, indicating that they are quirky subjects; consider this example with the verb hafa vantað “to lack,” which uses the auxiliary verb hafa “have” and takes an accusative subject:
- Svein hefur oft vantað peninga í vetur. • “SveinnACC often lacks money this winter.”
- →Í vetur hefur Svein oft vantað peninga. • “This winter, Sveinn often lacks money.”
The movement of í vetur “this winter” to the beginning of the sentence triggers inversion of the subject and the auxiliary hefur, even this accusative subject Svein.
Subject tests in Old Icelandic
Old Icelandic syntax behaved a bit differently from that of modern Icelandic, so some tests are more difficult to conduct. For example, reduction in coordinate structures is not a viable test because Old Icelandic regulary deleted subjects, objects, and other arguments when contextually clear. What follows are some examples of tests that do suggest quirky subjects in Old Icelandic
Subject-auxiliary inversion
Negation is another context which can trigger subject-auxiliary inversion. Here is an example from Svarfdæla saga with inversion of the accusative subject þig “youACC” and the auxiliary verb munu ”will, shall”; the verb which takes the accusative subject here is skorta “to lack”:
- Ei mun þig hér mat skorta. • “YouACC will not lack food here.”
Raising to object
The verb in the subject-to-object raising example from Njáls saga is vera nær skapi “to want” or, more literally “to be near [one’s] mood, which takes a dative subject:
- Gunnar sagði sér það vera nær skapi. • “Gunnar said that this was what heDAT wanted to do.”
The next example from Ljósvetninga saga uses the verb missýnast “to see wrongly,” which also takes a dative subject:
- Þórður… kvað Þorgeiri mjög missýnast… • “Þórður said ÞórgeirDAT must see wrongly.”
Conclusion
Many constructions in old & modern Icelandic morphologically match the structure of impersonal verbs in other languages, but their oblique (non-nominative) arguments behave as syntactic subjects. This behavior presents an interesting situation in which the morphology of case marking and the syntactic behavior of arguments conflict, and suggest that a “subject-spectrum” may exist in many languages which exhibit impersonal-verb behaviors.
bibliography
- Barðdal, J & Þ Eyþórsson 2018. “What is a Subject? The Nature and Validity of Subject Tests.” Ghent U & U of Iceland, Draft
- E Rögnvaldsson 1992. “Quirky Subjects in Old Icelandic.” U of Iceland
- Poole, E 2014. “Deconstructing quirky subjects.” North East Linguistic Society
- Þ Eyþórsson & Barðdal, J 2005. “Oblique Subjects: A Common Germanic Inheritance.” Language 81,4, p 824-81