How to Syntax 1 (the now that-Construction)

This is the first of a series of blog posts showing how I think about a syntax problem when I first notice it. I will occasionally choose phenomena that I notice, and talk about them in an informal fashion, breaking down the process of preliminary syntactic exploration. That is, I am just thinking off the top of my head (brainstorming), with few or no revisions. The focus of the discussion will be on process. I am not trying to come up with a polished analysis. Of course, if people suggest references for me to look at, I will look at them later, but that would be a second stage of thought, not the preliminary exploration.

Now That

In English, the following construction is used:

(1) Now that I have finished, I will leave.

I will call the phrase [now that I have finished], the now-phrase for short, and the construction the now that-construction. I have never read anything written about this construction in my career. I also did not try to look it up in the standard sources (e.g., JSTOR). Lastly, I have done no internet searches so far trying to confirm or disconfirm any of the data points below.

How did I find the construction? I have had it on the back burner for a few weeks, but I can't honestly say that I remember the exact moment I noticed it as being interesting. But the way I find most of my constructions is by a kind of monitoring/observing of the ambient language around me. There are periods when I monitor what I am saying or hearing or reading, and analyze it a bit. In that way, I stumble across lots of interesting things. Most of my papers on English syntax/semantics come from that kind of process.

Basic Combinatorics

The first stage in preliminary exploration is what can be called basic combinatorics (or alternatively, basic experimentation). Just try deleting, inserting, replacing and permuting various words and morphemes in the construction to get a feel for how it is put together.

For example, it seems the that is obligatory. Consider (2) (a case of deletion of a word):

(2) Now I have finished I will leave.

While (2) is good, it seems to have a structure very different from (1). Rather, (2) seems to be two separate sentences. So, the punctuation should be:

(3) Now, I have finished. I will leave.

But in (1), there seems to be only one sentence, not two. For example, consider the punctuation:

(4) Now that I have finished. I will leave.

The punctuation in (4) makes the now-phrase stand on its own as a sentence, but that is impossible. 

The generalization that that cannot be omitted in (1) makes one try to think of other constructions where that can and cannot be omitted. I call this process searching for syntactic parallels (or for related constructions). The search for syntactic parallels is a very important part of syntactic exploration and can lead to many different kinds of empirical questions and testable hypotheses. 

For example, in manner relative clauses and temporal relative clauses the that is freely omitted:

(5) a. The way (that) John treats me is unacceptable.

        b. The day (that) I saw John, it was raining.

But in other relative clause that cannot be omitted (e.g., subject relatives). So right away we have a question about how (1) is related to various relative clause constructions. 

Another combinatoric generalization is that the now-phrase can appear in different positions (permutation of phrases). In (1), it appears initially, in (6) it appears finally:

(6) I will leave now that I have finished.

But even though the now-phrase can appear either initially or finally, it cannot be broken up:

(7) *Now, I will leave, that I have finished.

So, the now-phrase seems to be constituent with an obligatory that. 

Other temporal words (different from now) do not seem to allow the same kind of construction (replacement of a word):

(8) a. *Tomorrow that I finish, I will leave.

b. *Yesterday that I finished, I left.

c. *When that you finish, you will leave.

d. *Then that I have finished, I will leave.

Even modification of now seems impossible (insertion of a word):

(9) a. Right now, I finished.

b. *Right now that I finished, I will leave.

c. Just now, I finished.

d. *Just now that I finished, I will leave.

And even synonyms do not allow the that-clause. Assuming now and at this moment are synonymous, consider:

(10) a. At this moment (= now), it is raining.

b. *At this moment (=now) that I have finished, I will leave.

Another basic combinatorial fact is that now and now-phrases do not always have the same syntactic distribution:

(11) a. Now is the best time to leave.

b. ??Now that I have finished is the best time to leave.

(11b) sounds very awkward. The generalization seems to be that now as a subject cannot be replaced by a now-phrase. The unacceptability of (11b) seems to be related to similar facts with subordinate adverbial clauses:

(12) *Since I have now finished is the best time to leave.

(12) constitutes a syntactic parallel between now-phrases and subordinate adverbial clauses.

Ultimately, a syntactic analysis will have to account for these basic combinatorial facts. In a final paper, it is not enough just to notice a bunch of stuff. But crucially, the analysis will be successful to the extent it provides explanations for these facts.

Paraphrase and Entailment

All the above facts bring up the issue of the syntactic and semantic relation between now and the that-clause in a now-phrase. Faced with this issue, the first thing is to try to find a paraphrase of (1) to illuminate the structure and meaning. This stage of investigation is called paraphrase and entailment.

Here is one attempt:

(13) Since I have finished, I will leave.

(13) has roughly the same interpretation as (1), but (13) seems to be lacking the temporal implication that (1) has. In other words, there seems to be the following implication:

(14) a. Now that I have finished, I will leave. (implies)

b. I have now finished.

The same implication does not seem to hold for (15) (similar remarks hold for given that):

(15) a. Since I have finished, I will leave. (does not imply)

b. I have now finished.

In other words, in (15a), maybe you finished a while ago, not now.

Here is another example illustrating the same fact:

(16) a. Now that I am over 18, I can vote. (implies)

b. I just turned 18.

(17) a. Since I am over 18, I can vote. (does not imply)

b. I just turned 18.

Here is another attempt at paraphrasing (1):

(18) Immediately after finishing, I will leave.

The problem here is that an inference like (19) still does not go through:

(19) a. Immediately after finishing, I will leave. (does not imply)

b. I finished now.

If one says (19a), there is no implication of finishing now.

One approach that seems to work better is the following:

(20) a. Since I have now finished, I will leave.

b. Given that I have now finished, I will leave. 

These seem very close to (1) in interpretation. The paraphrases in (20) are close enough to (1) that they provide preliminary hypotheses about what the structure of (1) is. 

Let me clarify that I am not claiming that paraphrases are in all cases the correct syntactic analysis. But used judiciously, they provide clues as to what the correct syntactic analysis may be. The basic assumption is that meaning/interpretation is derived from syntactic structure. So if the meanings/interpretations are the same, there might be a similar syntactic structure. This is not a logically rigorous conclusion, but rather a way to generate a testable hypothesis.

The basic idea is that now is moved to the front of the clause and something like since or given is silent:

(21) a. now SINCE I have <now> finished, I will leave.

b. now GIVEN that I have <now> finished, I will leave.

This analysis leaves open many questions. For example, are there other contexts where SINCE and GIVEN have null occurrences in English? Another question is this: If now-preposing triggers a silent SINCE or GIVEN, why can’t there be a silent SINCE or GIVEN with other temporal expressions:

(22) a. Since I finished yesterday, I will leave.

b. Given that I finished yesterday, I will leave.

(23) *Yesterday that I have finished, I will leave. 

An alternative analysis to the one in (21) is that (1) is some kind of relative clause construction where the head noun is now. Adopting the head raising analysis, the structure would be as in (24):

(24) now that I have <now> finished, I will leave.

But the same criticism leveled against (21) holds here: why can’t the construction work with other temporal expressions such as yesterday? 

Movement Diagnostics

Once we reach the level of a specific analysis as in (21) or (24), we can ask more detailed questions involving well know syntactic diagnostics. If now undergoes movement (as in (21) or (24)) does it show movement diagnostics? Can the movement be successive cyclic? Consider the following:

(25) Now that I said that John has finished, I will leave.

The question is what the interpretation of this sentence is. There are at least two possibilities:

(26) a. Since I have now said that John has finished, I will leave.

b. Since I have said that John has now finished, I will leave.

It is clear to me that the only interpretation of (25) is (26a), not (26b). This suggests that if now undergoes movement in the analysis of now that-constructions, it cannot undergo long-distances successive cyclic movement.

Conclusion

I have discussed some methods of preliminary syntactic exploration in this blog post. A summary of the methods is given here:

(27) a. Basic Combinatorics

b. Searching for Syntactic Parallels

c. Paraphrase and Entailment

d. Movement Diagnostics

This is only the very first step in cataloguing the methods syntacticians use in preliminary exploration. I hope to be able to post many similar blogs filling out this list in the future.

Acknowledgments: Thanks to Richard Kayne and Paul Postal for discussion of this construction.