Home » Articles » NG51. Attaching simple adverbs (1)

NG51. Attaching simple adverbs (1)

Truly...Attachment of adjuncts is a large and complex part of syntax.  This piece applies NG analysis to show how simple sentences can incorporate simple adverbs.  The results are good enough to encourage further work on more challenging structures.

The behaviour of particles was examined in NG24 to NG27.  Particles have some adverb-like characteristics and constrain the attachment of other material to the same verb. As particles can only occur after the verb, the discussion here starts with adverbs post-verb.  Each of the pairs below includes one sentence from those earlier pieces followed by a sentence substituting gladly for up.

Explaining adverb behaviour depends on NG’s change-of-C (category) mechanism as introduced in NG26.  The subscripts used here are context-specific, so giving2 in two different sentences doesn’t necessarily mean the same C.

Adverb but not particle

In sentence-final position, an adverb is possible but not a particle.

(5c) * Nero is giving Olivia to Poppaea up

(5f) Nero is giving Olivia to Poppaea gladly

(8c) * Nero is giving to Poppaea Olivia up

(8f) Nero is giving to Poppaea Olivia gladly

(17b) * Nero is giving to Poppaea up

(17e) Nero is giving to Poppaea gladly

(20c) * Nero is giving Poppaea Olivia up

(20f) Nero is giving Poppaea Olivia gladly

LS27 claimed that sentences like (5c) fail because the C (category) for giving changes at giving1__to or giving1__Poppaea, and there is no rule allowing the junction giving2__up.  To sustain that analysis and account for sentences like (5f), we must assume that the junction giving2__gladly is allowed.

Passives behave the same way even if they end with a by-phrase.

(11b) * Olivia is given to Poppaea up by Nero

(11e) Olivia is given to Poppaea gladly by Nero

(12b) * Poppaea is given Olivia up by Nero

(12e) Poppaea is given Olivia gladly by Nero

Particle but not adverb

In some other positions, a particle is possible but not an adverb.

(5a) Nero is giving up Olivia to Poppaea

(5d) * Nero is giving gladly Olivia to Poppaea

(7a) Nero is giving up Olivia

(7d) * Nero is giving gladly Olivia

The junction giving1__gladly is not allowed.  But giving1__up is possible and changes the C for giving such that giving2__(noun) and giving2__to are possible.

Neither

Right of indirect object Poppaea and left of direct object Olivia, neither particle nor adverb is possible.

(8b) * Nero is giving to Poppaea up Olivia

(8e) * Nero is giving to Poppaea gladly Olivia

(20b) * Nero is giving Poppaea up Olivia

(20e) * Nero is giving Poppaea gladly Olivia

Junctions giving1__to and giving1__(noun) are possible and change the C for giving such that giving2__up and giving2__gladly are not allowed.

The double-object form also fails with a particle or an adverb to the left of Poppaea.

(20a) * Nero is giving up Poppaea Olivia

(20d) * Nero is giving gladly Poppaea Olivia

LS27 analysed (20a) saying that giving1__up imposes a different C which allows an alternative rule for giving2__(noun).  This rule imposes yet another C which disallows any following giving3__(noun).  (20d) can be accounted for by assuming that giving1__gladly behaves in the same way as giving1__up.

Passives with GOAL as subject behave in essentially the same way.

(12a) * Poppaea is given up Olivia by Nero

(12d) * Poppaea is given gladly Olivia by Nero

In contrast, passives with THEME as subject – see (11a)/(11d) below – are grammatical because they have no following given__(noun) junctions.

Both

In the remaining sentences, a particle or an adverb can occupy the same position.

(5b) Nero is giving Olivia up to Poppaea

(5e) Nero is giving Olivia gladly to Poppaea

(7b) Nero is giving Olivia up

(7e) Nero is giving Olivia gladly

(8a) Nero is giving up to Poppaea Olivia

(8d) Nero is giving gladly to Poppaea Olivia

(11a) Olivia is given up to Poppaea by Nero

(11d) Olivia is given gladly to Poppaea by Nero

(17a) Nero is giving up to Poppaea

(17d) Nero is giving gladly to Poppaea

(19a) Olivia Nero is giving up

(19d) Olivia Nero is giving gladly

Sentences (8a)/(8d) are clunky but grammatical.  The analysis is simple.  The rules allow giving1__up and giving1__gladly.  The C doesn’t change to giving2 until giving1__to is subsequently processed.

Bare nouns following

The grammaticality of sentence (8d) is marginal.  Indeed classifying it under Both prevents the above evidence suggesting a general conclusion.  This would be an adverb following a verb prevents attachment to the same verb of a bare noun on the adverb’s right (although it does allow attachment of a preposition phrase).

The conclusion would be stronger than that for a particle, where a second bare noun following it is ungrammatical but not the first.  NG can handle this.  The discussion of (20a) under Neither would need to be elaborated with a second C being different for giving1__up and for giving1__gladly.

Shades of grey

Another point arising from (8d) is that the distinction between grammaticality and ungrammaticality is not so clear for an adverb as it is for a particle.  Why should that be?

Earlier LanguidSlog posts have speculated about the possibility of cognition, downstream from rule-based sentence processing, being able to make sense of ‘ungrammatical’ sentences.  The evidence for this includes our ability to deal with imperfect utterances by infants and non-native speakers, and with unfamiliar dialects.  Although unconscious, downstream processing is effortful compared with the fully automatic processing of sentences in our own dialect.

That being the case, we can speculate that an out-of-position gladly is easier to deal with in this way than an out-of-position up.  There is relevant semantic content in gladly, even in isolation.  In contrast, up in isolation doesn’t convey any of the semantics of GRUDGE or QUIT or YIELD as discussed in LS25; it’s more likely to suggest the locative concept of preposition up.  This could be why out-of-position up is more sharply ungrammatical.

Adverb plus particle

How do an adverb and a particle in the same sentence interact?

(5g) Nero is giving up Olivia to Poppaea gladly

(5h) Nero is giving up Olivia gladly to Poppaea

(5i) * Nero is giving up gladly Olivia to Poppaea

(5j) * Nero is giving gladly up Olivia to Poppaea

(5k) Nero is giving Olivia up to Poppaea gladly

(5l) Nero is giving Olivia up gladly to Poppaea

(5m) * Nero is giving Olivia gladly up to Poppaea

(5n) * Nero is giving gladly Olivia up to Poppaea

(11g) Olivia is given up to Poppaea by Nero gladly

(11h) Olivia is given up to Poppaea gladly by Nero

(11i) Olivia is given up gladly to Poppaea by Nero

(11j) * Olivia is given gladly up to Poppaea by Nero

The adverb cannot occur between verb and particle.  Again it is changes to the C for the verb that control this.  Evidently junctions like give1__up and give1__gladly are both allowed and both change give1 to give2.  Junction give2__gladly is possible but not give2__up.

More shades

Sentences (5m) and (11j) seem not too bad.  Is up being treated as a preposition with some sort of locative meaning and a PP complement?  That would be true for a variant of (5m) with a verb that doesn’t allow up as a particle.

(193) Nero is handing Olivia gladly up to Poppaea

But the reasoning is still not convincing because hand does allow particle over which, in this construction, better than (5m).

(194) Nero is handing Olivia gladly over to Poppaea

The choice is between abandoning a treatment of adverbs that makes (5m) and (11j) ungrammatical or contriving an explanation for why these sentences are not too bad.  The latter is preferred because finding an alternative explanation for post-verb adverbs seems unlikely.  The contrivance proposed is that automatic sentence processing does treat up (or perhaps up to) as a preposition and so these sentences are not sharply ungrammatical.  It is left to cognition to rearrange them. A native speaker can sense that something is going on because the sentences are certainly awkward.

Lots more to do

This piece has looked at only a few of the possible sentences with post-verb adverbs.  I hope this is sufficient to show that the change-of-C tactic described first in NG26 can easily account for gladly being grammatical in some positions but not in others.  What I’ll not attempt is to explain why those variations exist.

Comments