Tuesday afternoon online an acquaintance inquired, “In the idiom ‘mir ist langweilig,’ is there an implied (es)?”
I provided an answer with which I was not entirely happy. I also came to the realization, while untangling said answer, that the question was tied to a couple more general issues in linguistics (syntax in particular) and arguably based on some assumptions that reveal a good deal about linguistics, or at least common notions of the field.
In this post I will provide background to the specific question, go over my answers to my acquaintance, and in a rhapsodic rather than deeply structured manner hint at some of the broader linguistic, scientific, and philosophical issues raised by this specific question.
I. “mir ist langweilig.”
Is there an implied “es”?
First let’s provide some context for those who were not part of the conversation. The expression “mir ist langweilig” is German and roughly translates as “I’m bored.” A more literal translation would be “It’s boring to me” … except, no, that makes an assumption, that the underlying sentence is actually “mir ist es langweilig” or “es ist mir langweilig” (the ‘es’ being ‘it’).
A. German has four noun cases, four potential forms for nouns and pronouns — the nominative, accusative, dative, and genitive — that are used depending on a noun’s or pronoun’s function in an expression. The nominative is used for subjects, the accusative for direct objects, the dative for indirect objects, and the genitive for certain forms of possession or origin. But it’s not a one-to-one mapping. All cases have multiple functions (the nominative is also used for predicate nominative constructions, the accusative is also used for some time expressions), and a given function (such as being the object of a preposition) is often spread across multiple cases (accusative, dative, and genitive). The nominative is considered a “default” in a sense: it’s a starting point and we manipulate the nominative to get the other forms, so we like to think. When in doubt, use the nominative.
Something like that.
English has all the same functions as German, but only two forms: a nominative (I, he, she, we …) and an objective form (me, him, her, us …)
‘mir’ is the 1st person singular dative form, translated usually as ‘me’ or ‘to/for me’ in English.
B. The German verb ‘sein,’ ‘to be,’ is practically the only ‘irregular’ verb in the language.
German verbs fall into two large categories: weak and strong; these are not the same as regular and irregular, as both follow ‘regular’ rules, just different types of them. In general, strong verbs are old verbs that follow derivations of Indo-European ablaut; that is to say, as you conjugate the verb there is a prescribed progression of vowel changes (and accompanying consonant changes sometimes). We still have some of this in English. Consider ‘to sing’ … sing, sang, (have) sung. Contrasted with this are weak verbs, which perform no internal changes; they only change the ending of the verb according to person, number, and tense. Consider, for example, the English weak verb ‘to kill’ … kill, killed, (have) killed. There are also some verbs we call “mixed” and which some textbooks refer to as “irregular weak verbs.” They are weak because they follow the same pattern of verb endings, but irregular because they have some vowel changes; they are mixed for the same reason: they combine the traits of strong and weak verbs. An example is ‘to bring’ … bring, brought, (have) brought. The ‘t’ (usually a ‘d’ in English) in the past and participle forms is indicative of weak verbs, but the vowel change suggests strong.
In the verb ‘to be’ in English, German and other Indo-European languages we have something that fails to fall into one of the limited, established categories, something worthy of the label ‘irregular.’ This is not something necessary (‘essential’); it is a contingent relic of language history, of, for example, multiple verbs over time being combined into one abstraction that was then materialized. The reason why ‘be’ and ‘is’ and ‘was’ seem unrelated and not derivable from the other(s) is because they are (and are parts of different verbs). We do, however, get ‘been’ from ‘be,’ and ‘were’ from ‘was.’ German has ‘sein,’ and in the regular present tense such forms as ‘bin,’ ‘bist,’ ‘ist,’ ‘sind’ and ‘seid’ … only the last two appearing related to the infinitive. ‘ist,’ as in the construction that started all this reflection, means ‘is,’ and is the third person singular indicative, agreeing in person, case, and number with ‘er’, ‘sie,’ and ‘es’ (he, she, it). Not with the ‘mir’ (first person, singular, dative) that is provided.
C. ‘langweilig’ is just an adjective, one that means ‘boring.’ Two things are worth noting.
First is that it is a compound composed of ‘lange’ and ‘Weile,’ eventually becoming one word. It began as a noun construction meaning a long period of time, and would correspond to the English ‘long while.’ It eventually took on the meaning of ‘boring’ … I’m in a way reminded of an exchange in “Clue!” … Tim Curry begins, “To make a long story short …” and those around him shout, “Too late!”
Secondly, it is ‘boring,’ not ‘bored.’ ‘-ed’ adjectives in English often derive from verbs; they’re identical in form to past participles. And in German there is the verb ‘langweiligen,’ to cause boredom, the participle of which is ‘gelangweilt.’ In German the past participle can be used as an adjective, and that adjective can also be turned into a noun (usually as ‘the person or thing ______-ed’ or in this case, ‘der Gelangweilte’ would be the person who was made bored’), but most commonly show up in the passive voice.
Adjectives in German can take endings depending on the adjective’s function, of which there are two: attributive and predicate. An attributive adjective functions like ‘existence’ in the essence-existence pair, as the ‘part’ in the whole-part pair … as the ‘attribute’ in the object-attribute pair! Nouns are objects; adjectives are their attributes describing which one or what kind. The attributive adjective in German, as in English, comes before the noun. Without an adjective we have “That’s the man!” but with it we have “That’s the old man!” or “That’s the red-haired old man!” or similar. What’s great about this is that we realize we should treat anything, not just a single word, that comes before a noun and describes which one, what kind, or how many as an ‘adjective.’ German, for example, has what we call ‘extended adjectival modifiers.’ They function as adjectives (who many, which one, what kind) but are extended insofar as they consist of multiple words. It’s easy to read “red-haired” in “red-haired man” as just a multi-word adjective, but German allows not just multiple words, but also complex structures involving other phrase types, especially prepositional phrases
Consider “der auf dem boden liegende Hund,” which is literally “the on the floor lying dog,” but which in English we would render as “the dog lying on the floor” or “the dog that is lying on the floor.” The second version uses a relative clause, and German can do this, too: “der Hund, der auf dem Boden liegt.” The relative clause employs a subject (pronoun) and finite verb; the extended adjectival modifier uses a participle as an adjective and thus needs no extra subject (noun or pronoun).
Attributive adjectives in German take endings that agree in case, number, and gender with the nouns they modifier. This is a compact way of saying that depending on the noun, the adjective will have a different form. It’s “das weisse Haus” but “ein weisses Haus.” It’s “eine alte Frau” but “ein alter Mann,” and “der junge Mann” but “dem jungen Mann.” Unlike in most modern Romance languages adjective endings in German do not mirror or map in an easy one-to-one way to the endings of the nouns they modifier (compare: the restaurant chain “El Pollo Loco,” in which the ‘o’ in loco matches that in ‘pollo’).
In contrast we have predicate adjectives, which do not take adjective endings and which do not immediately precede a noun. In English we might say “She is smart” or “They are nice.” In German we would say “Sie ist klug” and “Sie sind nett.” Neither ‘klug’ nor ‘nett’ (smart/clever and nice, respectively) has an ending; both follow a verb of being in this case forms of ‘sein,’ to be). This structure parallels the predicate nominative construction, in which a noun in the nominative follows a verb of being, e.g. “Er ist Professor” (He is a professor). Predicate adjectives and nominatives ascribe a trait (adjective or noun) to a noun or pronoun; they are equivalences, and we might think of the verb linking the two ‘sides’ as an equal sign.
‘langweilig’ in the given expression takes no ending and does not precede a verb; we would therefore treat it as a predicate adjective.
II. “In the idiom ‘mir ist langweilig,’ is there an implied (es)?”
Literally the expression means “me is boring” and we take it to me “I’m bored.” Sometimes we rearrange it as “It’s boring to me” or “I’m bored by it,” with both of the latter two incorporating an “it,” a German “es.”
The question above asks, in the expression “mir ist langweilig” has an “es” been omitted?
“es ist mir langweilig” and “mir ist es langweilig” are two grammatical German sentences that mean the same thing as “mir is langweilig.” So was ‘es’ merely omitted? Was it ‘implied’ but just not stated?
A. It is generally accepted that you always need a subject with a finite verb in German. In fact this ‘need’ helps to explain so-called ‘dummy subjects’ in any number of expressions.
Consider the statements “die Sonne scheint” and “es regnet.” The former means “the sun is shining” and the latter “it is raining” (or “the sun shines” and “it rains”). Both refer to the weather, both are in the present tense, and so on. Both have subjects, “the sun” and “it,” but while on the surface they appear similar, we have to wonder, “I know it is the sun that is shining, but what is raining?” What is this ‘it’? What is the ‘it’ in “it’s snowing?” Is there some invisible, neuter sky giant up there urinating upon the world? Or is the case that we have an activity (raining) expressed by a verb (to snow) that we want in a finite form (present, not past, etc.), and so that verb needs a corresponding grammatical subject?
Why?
Because.
Isn’t that what makes a verb “finite”? An infinitive is unbounded, unlimited, because it doesn’t have a tense/time, a mood, a person or number; when we conjugate it and make it finite we assign tense, mood, person and number, and those attributes are reflected not just in the form of the verb but also in providing a subject.
Of course it does not have to be this way. Other languages can do it differently, and even in German there are instances of finite verbs without (explicit) subjects, most notably in imperatives. Even in English we tend to do this: we say “stand up!” not “stand you up!” Here the subject does not need to be explicit.
“mir ist langweilig” is equivalent to “es ist mir langweilig” or “mir ist es langweilig.” It makes sense to say, “the ‘es’ is implied.”
B. There’s something funny about all of this. There is no explicit subject in the expression, but finite verbs generally need explicit subjects, so what’s going on is that in the ‘actual expression’ (our ‘tokens’ in a way) we have no explicit subject, but the ‘es’ — subject — is implied, so “mir ist langweilig” *really* stands for “mir ist es langweilig.” But why the ‘es’? Because a sentence needs a subject, and so the ‘es’ is a dummy subject … one that doesn’t actually mean anything. In this more literal translation of the German, ‘it’s boring to me,’ what’s the ‘it’? What is this ‘it’ that is boring? Life? Space-time? Existence? Some unnamed situation? That’s what a neuter, singular pronoun gets us: a generic “whatever.”
That is to say, there is something ironic about claiming an implied subject when that (implied) subject is merely a placeholder, or, even better: a bug in the system.
The ‘easiest’ and ‘clearest’ answer is that there is an implied ‘es’ because you want ‘real’ sentences to have subjects, because finite verb require subjects (almost always) in German.
But given that there is little a priori reason to believe a sentence requires a subject (unless we arbitrarily define a sentence as consisting of a subject and finite verb, et.c), let alone an explicit one, we might as well consider counter-arguments
III. Other Considerations
We can look at alternatives from within and without German or the structures being investigated; we can also take a meta or methodological perspective.
A. Let’s assume for the moment that ‘mir ist langweilig’ is not (just) a reduced version of ‘mir ist es langweilig’ or ‘es ist mir langweilig,’ and instead let us just look at the form given us and look for comparable structures.
Consider, for example, the predicate adjective and predicate nominative constructions. The two are superficially parallel and both express a version of the statement “X = Y,” where ‘X’ is a noun or pronoun, ‘=’ is a verb like ‘to be,’ ‘to become,’ or ‘to remain,’ and ‘Y’ is a noun or adjective.
Consider, too, the German expression ‘mir gehört das Auto.’ In this case we have a dative pronoun (though it could be a noun, as in ‘Jürgen gehört das Auto’), a verb (in this case ‘gehören,’ to belong to) and a noun in the nominative (the subject, in this case ‘das Auto,’ the car).
It seems parallel to ‘mir ist langweilig,’ in which we have a dative pronoun (though it could be a noun, as in ‘Jürgen ist langweilig’ … oh, wait! I’ll have to remark upon this), a verb (in this case ‘sein,’ to be) and an adjective (‘langweilg,’ boring).
We might propose an analogy:
“mir ist [adjective]”:”mir gehört [noun]”::”ich bin [adjective]”:”ich bin [noun]”
That is to say, the ‘mir ist …’ construction appears similar to the ‘mir gehört …’ construction in a way similar to the predicative adjective vis-a-vis the predicate nominative.
There are, however, a number of problems with and objections to this, a couple of the mosts obvious being to me:
1. Predicate adjective and nominative constructions have the same rules, the same word order. What you can do with one you can do with the other. But in ‘mir ist langweilig’ vs. ‘mir gehört das Auto,’ the latter can be transformed into ‘das Auto gehört mir,’ but we do not transform the former into ‘langweilig ist mir.’
2. The predicate adjective and nominative constructions likewise do not care what comes in the ‘X’ position, but whereas ‘mir is langweilig’ and ‘mir gehört das Auto’ work in a parallel fashion, once it is no longer clearly a dative pronoun in the ‘X’ position but, rather, a noun, the structures differ: ‘Jürgen gehört das Auto’ is still ‘the car belongs to Jürgen,’ but ‘Jürgen ist langweilig’ is not ‘Jürgen is bored,’ but rather the predicate adjective construction ‘Jürgen is boring.’ To get ‘Jürgen is bored’ we need to add an ‘es’: ‘Jürgen ist es langweilig.’
The analogy is suggestive, but it falters if not fails upon generalization. Then again: that’s a trait of analogies in general.
Analogy and parallel structure also suggest that the ‘es’ is implied. If we are looking for another structure similar to ‘mir ist [adjective],’ instead of one that changes two elements (both adjective to noun, and the verb), let’s change only the verb. In addition to ‘dative + to be + adjective,’ German uses ‘dative + to go + adjective’:
(a) How’s it going? Wie geht’s / wie geht es (dir)?
(b) It’s going fine. Es geht mir gut; mir geht’s gut.
In this other construction the ‘es’ is not optional: ‘mir geht gut’ is not acceptable. The ‘es,’ however, is often elided: ‘mir geht es’ becomes ‘mir geht’s,’ and ‘wie geht es’ is ‘wie geht’s.’
B. One the questions raised was, why should we assume that finite verbs require a (nominative) subject?
Imperative constructions often elide the subject, which is nominative, but there are plenty of other examples of constructions that do not fit within our usual subject, direct object, indirect object framework.
1. Consider ‘dative objects’ and ‘dative verbs’ in German.
The nominative case governs subjects, the accusative direct objects, and the dative indirect objects. But there are a number of verbs, labeled ‘dative verbs’ in most textbooks, that do not take an accusative direct object; they only take objects in the dative. But these are not, precisely, indirect objects, but rather a kind of direct objects. Verbs that take direct objects are transitive, and transitive verbs can show up in both the active and passive voice; German dative verbs can appear in the passive, suggesting they are transitive and their (dative) objects are transitive direct objects.
In such instances we’re perfectly happy with a dative but no accusative object. By extension, why couldn’t we be happy with a dative but no nominative? Clearly the dative is not functioning as the grammatical subject of the sentence (as it is ‘mir ist,’ not ‘mir bin’ (which would cause at least person-number, but not case, agreement between ‘mir’ and the verb)).
2. Or consider the ergative.
The case of the noun used for the subject changes based on the kind of verb considered. Transitive verbs take the ergative, which marks an ‘agent’ of a verb; intransitive verbs take an ‘absolutive’ case, one that (like the nominative in English or German, is considered ‘unmarked’; it’s the ergative that *changes*); in languages with only an ergative and absolutive, the latter would be used for everything but the agents of verbs. Nez Perce, for example, has an ergative, accusative, and absolutive.
3. And there are other options as well, such as the antipassive.
In the active voice of transitive verbs the grammatical subject is the agent, the one who ‘performs’ the verb. In the passive voice the subject-agent is deleted and the direct object (accusative in the active) is promoted to nominative grammatical subject. What if we ‘shift’ once more and delete the semantic direct object and promote the indirect object (dative in the active voice) to nominative grammatical subject?
While not a feature of English, this is a common enough structure in Australian Aboriginal and Native American languages. The grammatical subject of a sentence that agrees with a verb need not be the agent (see: active, passive, middle voices). And the ergative reminds us that not all subjects are the same: in English and German we conflate the subjects of transitive and intransitive verbs, even though they can hardly be said to ‘mean’ the same or perform the same function.
But I digress.
In German if you transform a sentence that has only a dative object, not an accusative direct object, into a passive construction, there is nothing to automatically fill the role of a nominative grammatical subject. Consider the following example, ‘borrowed’ from Zorach and Melin’s “English Grammar for Students of German” (5th edition, p. 166):
(a) Man dankt ihm.
(b) Ihm wird gedankt.
(c) Es wird ihm gedankt.
The third (c) contains “es” as a dummy subject, but the impersonal “es” is not required, and we can see how (b) is derived from (a) as any other passive would be; the original subject is deleted and the accusative object is promoted to nominative subject, except here that position is blank. Here the dative object of the first sentence does not become the grammatical subject of the second (suggesting that ‘dative objects’ are not ‘direct objects,’ contrasted with what I claimed above). Compare:
(d) Ich esse den Apfel. (I eat the apple.)
(e) Der Apfel wird gegessen. (The apple is eaten.)
Here the accusative ‘den Apfel’ (direct object) becomes the nominative ‘der Apfel’ (subject).
Beth Levin, from the intro to her her summer ’05 course “Semantic Prominence and Argument Realization,” (part I, The Problem of Argument Realization) notes that “Dative/accusative experiencers in German impersonal constructions are subjects in English” and provides two examples taken from J. A. Hawkins’ “A Comparative Typology of English and German: Unifying the Contrasts,” (University of Texas Press: Austin, 1985. p. 56):
(f) Mich friert. ‘me freezes’ –> I’m freezing.
(g) Mir ist warm. ‘to-me is warm’ –> I’m warm.
There is no implied ‘es’ suggested here, though ‘mich friert’s’ or ‘mich friert es’ is also grammatical. The related ‘mir ist kalt’ (I’m cold) example is employed by Barry J. Blake in “Relational Grammar” (p. 114). But beyond simple ‘mir ist …’ type expressions or dative verbs put in the passive, there are other German constructions, many covered by Breckenridge in her 1975 B.A. thesis, that do not have explicit grammatical subjects, such as:
(h) Ihm wurde von jemandem im Raum gedankt.
(i) In MIT wird sogar am Sonntag gearbeitet.
(k) Ihnen war schwer zu helfen.
(k) Hier ist schwer zu tanzen.
The first two resemble the passive (b) above; but for (j) and (k) Breckenridge does not suggest a parallel to versions that have a dummy subject (though we might add a dummy ‘es’ after ‘war’ and ‘ist’ respectively) but to sentences featuring ‘object-raising’ (see the Introduction to “Rules Which Nothing Undergoes: An investigation of impersonal passives and object raising constructions in German” by Janet Breckenridge, aka Janet Breckenridge Pierrehumbert, 1975). Compare/constrast this with “German-English Verb Valency: A Contrastive Analysis” by Klaus Fischer” (e.g. p. 284 and subject-to-object raising), which also critiques Hawkins (referenced above).
In short, there are plenty of instances in German of there not being an explicit subject (even though a dummy subject may be inserted), and these often deal not with regular ‘subject-object’ language or matters of ‘nominative, accusative, dative,’ but rather of voice (active, passive … middle?), ergativity, or such roles as agent, undergoer, actor, patient, etc.
C. Methodological Notes
One assumption of the entire question and analysis is that there is a sense in which ‘mir ist langweilig’ and ‘es ist mir …’ or ‘mir ist es …’ represent the ‘same’ sentence.
In a slightly stronger sense, we are assuming there is a ‘deep structure’ corresponding to ‘mir ist langweilig’ and that the same deep structure corresponds to ‘mir ist es langweilig.’ Deep structure is a theoretical construct and it relates to ‘surface structures’ by unifying and explaining them. In a transformational-generative grammar operations on deep structure(s) allow us to convert, let’s say, ‘I love you’ into ‘You are loved by me.’
Deep structure is part of Chomsky’s project, implicitly and explicitly, from the late 50s through the 60s, but then we have “The Linguistics Wars.” We have Chomsky’s syntactic approach, which attempts to eschew semantics, and then we have other camps the create ‘generative semantics’ and so on … and at the same time either deep structure was located in deep semantics (rather than syntax) or deep structure was abandoned altogether. For that consider Randy Harris’ “The Linguistics Wars” (p. 128):
We have just ben looking at arguments that deep structure only has a few essential categories, that symbolic logic is the language of deep structure, that there is a universal base which implies all languages share a common core of deep structures, that deep structures call for absstract verbs, including abstract performatives, that deep structure is very deep indeed. And the next step is the abandonment of deep structure.
The signal document, the birth of announcement, the first important gauntlet, of generative semantics emerging from abstract syntax is a slight mimeographed paper — a letter, really – by Lakoff and Ross, entitled “Is Deep Structure Necessary?” (1976[1967]) which says no.
Perhaps we can back off from a Chomskian notion of deep structure, of one that commits us to greater pronouncements. Perhaps we just want to say that there is a useful level of abstraction that allows us to say these two sentence formulations are the same, or that one is derived from the other. But what is our authority for doing so? Is it only a useful heuristic? If so, against what are we judging this, and what would it even mean if it has no prescriptive force behind it? A purely pragmatic approach might just claim that by arriving at this abstraction we can treat the two (or the three) the same an analyze them the same way, ignoring minor differences or combining analyses of one with the other to make more far-reaching conclusions. If it is an ‘abstraction,’ it is about taking away superfluous details to focus on what is either essential and/or useful, the the former makes an assumption that there is something essential, and the latter becomes a matter of mere pragmatic relativism … we no longer say anything meaningful about our object of study, only what it’s good for in a certain, changing context. The charge of essentialism, though leads in a couple directions, one which I probably won’t have time to address in this piece, Platonism, and the other which is posed by the question: how do we know there is a deep structure, and even if there is, do we really have access to it?
There certainly can be a positive, scientific answer: we posit deep structure and hypothesize about it; then we make predictions; and after devising ways to test these predictions if they pan out, we’ve got a degree of confirmation. On the other hand we have to be careful because ‘deep structure’ is a bit like a Kantian ‘Ding an sich’ to which we have no direct access, only inferences. If we only have access to surface form (which I prefer over surface structure, at least as contrasted with deep structure), why even posit something deeper?
We tend to take that ‘deeper’ thing as ‘more real’ in an ontological sense because we take it as ‘primal’ in some sense, or first. In fact, much discussion of linguistic deep structure seems to replicate this causality of ‘firstness’ and our utterances and surface form ‘come from’ deep structure. This is the more strongly Platonic version, though I’ll leave that aside for now; a much more neutral though still somewhat loaded Kantian version would treat deep structure akin to versions of the related ‘universal grammar’ and less a matter of something positive, actual rules, let’s say, than constraints or, better yet, mere preconditions that would have to first be met. But recall that this deep structure is inferred from something we do have access to (surface form), so that in a way it (deep structure) is merely an ‘illusion,’ even though we tend to take it for ‘more real’ than the our empirical material.
It’s this paradox (and dialectic) of sorts that also links us to Platonism, to the allegory of the cave, and to idealism and rationalism.
What got us to his philosophical aside was mention of ‘causality,’ perhaps as something temporal: which came ‘first’? The empirically attested surface form or the theoretical deep structure? The latter has explanatory power; in that sense it ’causes’ the former (in the sense of Aristotelean ‘formal cause,’ perhaps). But the greater issue for us is that the moment we talk about deep structure we’re associating (not improperly) with transformational-generative grammar, with syntax, with a whole range of synchronic linguistics, and it’s in synchronic linguistics that we find rationalism at home, we find systems and rules. But whereas it provides a great ‘snap shot’ of a language at a given moment, it has a harder time across time, in diachronic or historical linguistics. The deep structures that may elegantly explain Modern English may fit quite poorly with Old English, though the former is clearly derived from the latter. See also: “Historical Linguistics” by Theodora Bynon (e.g. p. 155) for precisely this issue; that it is a problem is not to say it does not admit to solutions.
IV. Conclusions and Reflections
When asked “In the idiom ‘mir ist langweilig,’ is there an implied (es)?” my first response was “more or less.” A good German “jein” would have done the job. As a language teacher I’ve often suggested to students, assume there’s a dummy subject here. Assume finite verbs require grammatical subjects, so either include a ‘dummy’ (e.g. ‘es’) or assume it has been omitted colloquially but you and ‘read it in.’ At the same time this is not a real answer the question as posed; it is instrumental and pragmatic, it is useful for the language learner who comes from English and who has been taught to expect subject-verb agreement (even if — especially if?! — they do not practice it in all their own utterances). Perhaps it is a matter of falling back upon grammar (and grammar-translation) in the classroom; would it be an issue in a truly communicative setting?
So one answer was, “It’s useful to say so.”
But then I looked at it and could contemplate similar — some more so, some less so — constructions and ponder possible scenarios in which ‘mir ist langweilig’ was not a cognate of an ‘es’ construction but of some other structure. At the same time that the expression, as provided, was not very productive compared to, say, predicate adjectives, but it was parallel to plenty of other constructions in which an ‘es’ could be inserted.
Still, by refusing to simply accept that it was a derived form (on in which the ‘es’ was dropped, that the ‘es’ was optional) I had to consider other but related categories, such as ergativity, object-raising, antipassivity and the like. And there was always that weird feeling that it’s weird to think of the base form as one containing a dummy subject, since the whole point of a dummy subject was to provide something syntactically required but semantically vapid, yet if it could be dropped, then clearly it was not required at a deep level …
A. This stops here. It’s merely an exercise. But it’s also an illustration. It’s fascinating that one little expression, ‘mir ist langweilig,’ can tie into so many different issues.
B. This is just a rhapsodic aside. While contemplating this question when it was posed yesterday it reminded me of the work of a (former) colleague with whom I had a productive discussion one evening that reduced more or less to the matter of what direction does the causality flow … surface to deep or deep to surface? How were we going to explain syntactic change over time?
But that’s another question for another day.
