3

For those who came in late, a "sentence adverb" is a word that modifies an entire sentence rather than just the verb or predicate. A sentence adverb communicates speaker attitudes about the proposition that the sentence denotes, or discourse information. So "fortunately" is a sentence adverb in the sentence "Fortunately, we found a source of fresh water on the island." Also note "however," the sentence adverb for discourse.

I couldn't find much information on the Internet about sentence adverbs. Here are the three most helpful links I found:

http://www-01.sil.org/linguistics/GlossaryOfLinguisticTerms/WhatIsASentenceAdverb.htm http://grammar.about.com/od/grammarfaq/f/sentadvqa.htm http://www.oxforddictionaries.com/definition/english/sentence-adverb

All of the links define sentence adverbs as sentence modifiers. But is this always true? Note the following:

a) The man fell from the tenth story, fortunately onto a pile of pillows.

b) Fortunately, the man fell from the tenth story onto a pile of pillows.

In sentence a), "fortunately" clearly conveys speaker attitude and has the same meaning as it does in sentence b). However, the sentences don't necessarily have synonymous readings.

It can be unfortunate that the man fell from the tenth story onto a pile of pillows, and at the same time fortunate that he landed on a pile of pillows (rather than the hard pavement, for example).

So, if "fortunately" in sentence a) can convey speaker attitude about the state of affairs denoted by the prepositional phrase "onto a pile of pillows," what should we call "fortunately" in this context if not a sentence adverb?

I am assuming that other languages have sentence adverbs, and that this question is therefore not off-topic.

James Grossmann
  • 8,730
  • 8
  • 41
  • 83

2 Answers2

3

Adverbs have long been called a ‘wastebasket’ category in syntax. Their definition is very general: adverbs are distinguished from adjectives, which modify nouns, by saying that ‘adverbs modify verbs, adjectives, or other adverbs’; to this one can add that they may also modify phrases and clauses as a whole.

If something doesn’t fit nicely in some other word class, it generally gets lumped in as an adverb of some sort. Traditionally, adverbs (and adverbials, which means ‘any chunk that acts like an adverb’ -– not a terribly precise definition) represent qualifications and afterthoughts to ordinary propositions.

And they can occur in many different positions in the sentence, which means they may or may not be close to whatever constituents they focus on. Which allows lots of useful variations and conventions.

"Sentence adverb" is not a type of adverb.
It's a type of adverbial construction, or usage;
one of the things some adverbs can do.
That's all.

And "sentence adverb" is a sloppy technical term, since it doesn't distinguish "sentence adverbs"
from "clause adverbs", or "phrase adverbs", which also exist, but are not mentioned.
So I wouldn't worry too much about "sentence adverbs".

jlawler
  • 10,042
  • 1
  • 31
  • 55
  • I think we basically come to the same conclusion: it's not a very useful label for a word, but merely for a construction. What I tried was going by what are commonly called sentence adverbs and see whether I could make anything consistent out of them (and I couldn't). – Cerberus Oct 29 '13 at 18:07
  • Yeah, I kinda thought that was what you intended. That's why I don't think it's worthwhile to pay attention to "what are commonly called X", where X is any grammatical term at all. There simply isn't any consistency to be drawn from commonality, because there is so much BS in the system. It's like discussing comparative religion in Syria at the moment. – jlawler Oct 29 '13 at 18:25
  • Don't you think "describes the attitude of the speaker towards his statement" is a common element? – Cerberus Oct 29 '13 at 18:59
  • It's a totally ubiquitous element. Every word uttered "describes the attitude of the speaker towards his statement". It's like saying it "is audible to the addressee" or "conveys a meaning". Since it's true of everything, it doesn't serve to distinguish anything special unless "attitude" is pinned down much more clearly in semantic and pragmatic terms. – jlawler Oct 29 '13 at 20:44
  • I suggested modal (fortunately) and discursive (however) categories in my answer. I think those describe the attitude of a speaking in the examples given in a special and direct way that other adverbs do not. – Cerberus Oct 29 '13 at 23:50
  • Modality has a technical definition involving either Possible or Necessary; judgements about good results (fortunately) aren't really "modal". But and however both have presuppositions of unexpectedness, though the details are all contextual; I don't see how calling them discursive helps, since I don't know what it's sposta refer to. – jlawler Oct 30 '13 at 17:18
  • Modality comprises desirability and possibility. That's why I suggested that adverbs like fortunately, when expressing desirability from the perspective of the author, could be called modal. Discursive is commonly used to refer to words that explicitly indicate antithesis, importance, expectation, etc. from the perspective of the author—in short, ordering arguments/elements within and with respect to a discourse. As in discourse particles and discourse markers. – Cerberus Oct 31 '13 at 03:55
  • Desirability is usually thought of as only one deontic sense of a modal; the deontic sense of will, in fact -- the one that's required in hypothetical classes. But that's only one of the several deontic interpretations of modals. And, once again, how does one fashion tests to see whether given lexical items are or are not "words that explicitly indicate antithesis, importance, expectation, etc"? Not that your description is wrong, but it's not categoric and it's contingent. It's an effect or a goal, rather than a description. – jlawler Oct 31 '13 at 04:22
2

Good question. Your examples are convincing. As I see it, there are four possibilities:

  1. Fortunately is not a real sentence adverb; real sentence adverbs do modify the entire sentence.
  2. Fortunately does in fact modify a sentence, but this is obscured by ellipsis.
  3. The same adverb can be a sentence adverb in some sentences but not others.
  4. The term sentence adverb does not adequately describe the category of adverbs we have in mind.

1) Does fortunately happen to be a bad example? Let's take a few other, similar words, then.

a) The cat became agitated. However, the mouse was absent from the kitchen.

b) The cat was in the kitchen. The mouse, however, was not.

c) I bought a pair of hopefully unused socks at the market.

Here it can be seen that however modifies the entire sentence in a), but only the mouse in b) (contrasting it with the cat). And hopefully clearly modifies only unused in c), which is perhaps a bit informal, but not uncommon.

2) One could say your example is really an elliptical form of a full secondary sentence:

a) The man fell from the tenth story; fortunately [it was / he fell] onto a pile of pillows.

I find this doubtful.

b) The cat was in the kitchen. The mouse, however, was not.

It seems impossible to reconstruct this as ellipsis.

c) I bought a pair of — hopefully [they are] unused — socks at the market.

This perhaps possible, but it seems weak. And it cannot work for however. It seems far more likely that the subconscious sometimes wants to treat hopefully like other -ly words and have it modify an adjective. I hear an English monk sharpening his razor.

3) Perhaps the same adverb can sometimes be a sentence adverb, but at other times modify only a single constituent. This would then have to apply to most sentence adverbs. The problem or issue is that most adverbs that we normally do not consider sentence adverbs can also do this:

d) The mouse was often absent from the kitchen.

e) The often absent mouse was dearly missed by the cat.

What makes this different from fortunately and however? The main difference seems to be that the latter are about the attitude of the speaker towards the text, modality, whereas often is not. In the case of fortunately, its meaning does not change between its use as a sentence modifier and its use as modifying a constituent. It would not seem warranted to call the word a sentence adverb in one sentence and a normal adverb in another, if its meaning does not change; then we could just as well call often a sentence modifier in d), which seems trivial.

4) It doesn't seem to make sense to use sentence adverb at all, because it suggests that certain adverbs modify only whole sentences and other adverbs do not. Instead, the essence of what makes fortunately etc. special is that they express the attitude of the speaker towards the text, somewhat like modal verbs. Perhaps the intended category should be called modal adverbs instead. Or perhaps we are dealing with more than one category: however could be called a discursive adverb, as it does not really express a modal attitude, but rather the speaker's attitude towards the ordering of the text (in casu a contrast). An attitudinal adverb?

Note that the word disjunct has similar problems, because it implies that the adverb is entirely separate from the rest of the clause and does not modify a constituent.

TKR
  • 10,893
  • 26
  • 60
Cerberus
  • 7,976
  • 33
  • 49
  • 1
    I don't think it makes sense to say that in (b) however modifies only the mouse: an adverb can only modify a predication, not a single referent by itself. It seems clear to me that (despite its position) however in that example is modifying the entire sentence The mouse was not (in the kitchen). (Which makes the ellipsis question irrelevant for that example, as there's already a finite verb.) – TKR Oct 29 '13 at 16:20
  • I agree. However virtually always modifies a clause, though the clause may be deleted, because syntax. And I don't think these are the only 4 possibilities. Any more than I think "sentence adverb" is anything more than a nonce term for some particular uses of certain adverbs. There are lots of others. – jlawler Oct 29 '13 at 16:48
  • Tim Stowell has done some work recently on sentences like your 2(a), which he calls 'parenthetical adjuncts' - an ellipsis analysis is argued for, i.e. The man fell from the 4th story, fortunately [onto a pile of pillows]i <he fell from the 4th story ti>, where [onto a pile of pillows] is a focused adverbial which moves to the left-periphery of the adjunct, thus licensing clausal ellipsis. Unfortunately there isn't anything online (this was a talk from last week), but i found the arguments convincing enough. – P Elliott Oct 29 '13 at 17:11
  • 1
    Also @jlawler "...though the clause may be deleted, because syntax." <-- i love this comment. – P Elliott Oct 29 '13 at 17:12
  • @PElliott In that analysis, how do you get from two sentences (The man fell from the fourth story. Fortunately he fell from the fourth story onto a pile of pillows) to one? You can't generally just glom two sentences into one without a coordinator or subordinator of some kind, can you? – TKR Oct 29 '13 at 17:58
  • @TKR: Perhaps "modifies" is not right. But it seems clear to me that the position of however specifically marks the preceding constituent "the mouse" as contrastive. It is not the same as putting however at the beginning or at the end. There is some relationship between the position of however and this focus/contrast on a specific constituent. But it seems to be only possible with a constituent in first position. What else to call it? – Cerberus Oct 29 '13 at 18:02
  • @PElliott: I don't know, one can explain anything through ellipsis if you supple a long and complex enough phrase! And how would you propose ellipsis in a hopefully unused shirt? I'm not a huge fan. – Cerberus Oct 29 '13 at 18:04
  • However is a subordinating version of but. But, for its part, is and with a presupposition of false expectation. In almost any case, a subordinate however clause may be replaced with a conjoined but clause, salva veritate. Of course, coordinate clauses don't behave syntactically like adverbs. – jlawler Oct 29 '13 at 18:18
  • As far as the name goes, "ellipsis" is way too general; I'd just call it conjunction reduction. The man fell from the tenth story, (but) fortunately [the man fell from the tenth story] onto a pile of pillows. Nothing special here, looks like. – jlawler Oct 29 '13 at 18:23
  • 1
    @jlawler: Why do you call however subordinating (or even a conjunction)? I don't see The mouse, however, was not in the kitchen. as a subordinate clause, but as an independent sentence. – Cerberus Oct 29 '13 at 18:58
  • @Cerberus It's certainly true that The mouse is a contrastive topic in that sentence, but as far as I can see, that's the case whatever the position of however. – TKR Oct 29 '13 at 19:50
  • Many reduction rules work between adjacent sentences. There are no periods in speech, after all; only a flow of sound. Basically it's a memory problem; anything that's been referred to long enough ago becomes context, while anything that's been referred to recently becomes coreference. And the balance varies. Frequently. – jlawler Oct 29 '13 at 20:40
  • @TKR: I wouldn't say that: if you end or begin with however, the whole sentence can be contrastive, whereas putting however after the first constituent seems to mark it specifically. Just as putting not in a different position can have a different effect, even though the general meaning of the sentence doesn't necessarily change. – Cerberus Oct 29 '13 at 23:46
  • @Cerberus: The whole sentence is contrastive in any case: there's a contrast in predications, not just a contrast in subjects. However doesn't necessarily mark contrast specifically on what precedes it. Consider The cat pounced; the mouse, however, got away in time: here obviously the contrast is just as much about the actions in the VPs as about the referents of the subject NPs. In fact, if second-position however was about subject contrast specifically, you might expect the following to be grammatical, since only the subjects contrast: ?The cat ran away; the mouse, however, did too. – TKR Oct 30 '13 at 00:40
  • @TKR: I see what you mean: the other element in that sentence (got away) is also highly contrastive or "unexpected". But how about this. The dog was searching for the cat outside. In the kitchen, however, the cat was eating the dog's food. There is additional focus on in the kitchen. The main contrast is between outside and in the kitchen. The alternative is this: the dog was searching for the cat outside. However, in the kitchen the cat was eating the dog's food. The main contrast can be anything here. It could be searching for the cat v. eating the dog's food. – Cerberus Oct 30 '13 at 01:26
  • @Cerberus: I don't quite share your intuition about the 2nd example; in the kitchen seems contrastive there too due to its initial position. (I'm not sure what a 'main' contrast is, or how to distinguish it from other contrasts, unless by degree of stress.) I think it's just that contrastive topics ('focus' isn't the right term here) get placed initially, and however often occurs in the position following them. Maybe the reason these seem 'more contrastive' is that the position of however marks what precedes it as a prosodic unit, drawing attention to its function as a contrastive topic. – TKR Oct 30 '13 at 01:58
  • @Cerberus There's good evidence that this really does involve ellipsis. You get case connectivity effects in e.g. German: "Lena hat jemanden gestern gekuesst, hoffentlich einen schoenen Mann" (L. has someone.ACC yesterday kissed, hopefully a handsome.ACC man). The 'connectivity effect' is the fact that the putative ellipsis remnant following 'hopefully' is assigned the same case as its correlate 'jemanden' (accusative case). Where does the remnant get its case from? Presumably from inside the elided clause. – P Elliott Oct 30 '13 at 23:29
  • @Cerberus I agree with you that the hopefully unused shirt example doesn't involve ellipsis. There's no need to posit any elided material to derive the interpretation. That doesn't count as an argument against parenthetical adjuncts involving ellipsis. – P Elliott Oct 30 '13 at 23:30
  • @PEll: But why do you consider that evidence of ellipsis? Matching cases in apposition is normal, even the only option, in languages that have cases. I rather see it as evidence of how nouns and adjectives were not very different in Proto-Indo-European, in that any nomen could be used adjectivally or substantively. In Latin, you can turn any adjective into a noun, and many nouns into adjectives, like natio amica "a friendly tribe". Or Mars victor "victorious Mars / Mars, the victor", Venus victrix "victorious Venus / Venus, the victrix": it doesn't matter which way you translate it. – Cerberus Oct 31 '13 at 04:08
  • @PElliott: Right, perhaps the parenthetical construction is elliptical. Though I'm still not entirely convinced. And it only explains away a certain type of fortunately, so "sentence adverb" remains weak, in my opinion. – Cerberus Oct 31 '13 at 04:11
  • @Cerberus The case-matching observation isn't about the case on the adjective matching the noun, it's about the fact that the case on the entire Noun Phrase following 'hopefully' has to match the case of the indefinite correlate 'someone'. – P Elliott Oct 31 '13 at 13:21
  • @PElliott: I know that, and I don't understand why you see that as evidence of ellipsis rather than ordinary apposition. – Cerberus Oct 31 '13 at 14:14
  • If it's ellipsis, we have an explanation for why case-matching holds. I don't know what you mean by 'ordinary apposition', or how that would provide an explanation. – P Elliott Nov 01 '13 at 02:53
  • @PElliott: Cases have to agree in any apposition of two nouns/pronouns. Are you saying this is a special kind of apposition, or are you saying all appositions are elliptical? – Cerberus Nov 01 '13 at 12:23
  • @Usually It doesn't look like apposition to me - the nouns that match are completely discontinuous. Can you give me an example if what you mean? – P Elliott Nov 01 '13 at 12:45
  • @PEl: I just mean any kind of apposition, as in my example above Venus victrix, or ho Sokratês ho philosophos ("the Socrates the philosopher"), my father King Henry VIII, etc. Why do you consider your German example different, and how do you define discontinuous? (As in my comment above, I don't feel concord between two nouns is any different than between noun and adjective; historically, I don't think there was a huge difference between nouns and adjectives in the proto-languages. Most nomina could be used adjectivally and substatively.) Perhaps I should ask a Question about this... – Cerberus Nov 01 '13 at 14:39
  • Concord is usually extremely local, whereas in my German example, one of the nouns is an object in the matrix clause, and the other is embedded under a sentential adjunct. You'd have to allow concord to be extremely unrestricted to derive the matching facts, and i don't think that would be such a good idea. – P Elliott Nov 01 '13 at 15:14
  • @PElliott: I don't know what all that means, but this is just how it works in languages with cases; it doesn't seem remarkable or different to me. I don't think I would want a more complicated theory behind it than "it is in apposition to the other noun, so it gets the same case". It happens all the time in e.g. Latin, and I have never heard ellipsis suggested. – Cerberus Nov 01 '13 at 16:39