Li7 Hi sto ri cal li n gui stic s

Ch a nge a nd r ea n aly si s
http: //www.l i ng.ca m. ac. u k/li7 /

1 WHAT’S THE QUESTION? ACTUATION, ACTUALISATION AND TRANSMISSION

1.1 Some historical data: English stress alternations

Stress shift rule in verbs related to a noun:

(1) I dispute that.

(2) Mary took no part in the dispute.

(3) The default file on my computer just got deleted.

(4) Iceland may default on its loans again this year.

(5) She's one of the most direct people you'll ever meet.
(6) The police had to direct the traffic because of the accident.

VERB NOUN but: VERB NOUN

pervért pérvert cónvoy cónvoy
recórd récord préface préface
transpórt tránsport
convért cónvert
presént présent
contráct cóntract

It is (logically) possible to distinguish three stages / processes:

1. the rule appears in the language: Form a verb by moving the stress of the
corresponding noun from the first to the second syllable = actuation

2. the rule applies to more and more words in the language (or more and more
linguistic contexts) = actualisation / extension

1 DWW
3. more and more speakers (different social groups) more and more registers
(stylistic levels) = transmission (diffusion)

Many changes can be thought of in this way (or partly in this way).

Different linguists have different views on where ‘true’ change lies:

“The problem is that ‘one swallow doth not a summer make,’ and one
change in the grammar of an individual does not constitute what we
think of as a change in ‘a language.’” (Hopper & Traugott 1993: 36)

“Methodologically it is certainly preferable to recognize change only

when it has spread from the individual to a group…” (Hopper &
Traugott 1993: 38)

“…a general theory of change must provide an account for any

change, regardless of whether or not it diffuses.” (Hale 1998: 6)

2 WHY DOES LANGUAGE CHANGE?

Some general types of explanation:

2.1 Typological explanations

Changes are movements towards unmarked types of language e.g. languages
with symmetrical consonant systems, languages where modifiers all go on the
same side of the head.

Old English f p t θ s ʃ
b d

Present-day English f p t θ s ʃ
v b d ð z ʒ

2.2 External factors

• loans
• borrowing of sounds and constructions
• convergence

2.3 Generative (acquisition-based) model

Generative view: there is a discontinuity of transmission between each generation
and the next. Each generation has to create the rules of the language afresh.
Language change happens in this discontinuity by abduction (in particular during
child language acquisition) (Andersen 1973). Children’s mistakes today are
tomorrow’s linguistic changes.

2 DWW
 Grammar 1 Grammar 2
(Adult) (Child)

UNIVERSAL
GRAMMAR

Output 1 Output 2

DEDUCTION INDUCTION
LAW all men are mortal CASES Socrates is a man
CASE Socrates is a man Aristotle is a man
RESULT Socrates is mortal Plato is a man
RESULTS Socrates is mortal
Aristotle is mortal
Plato is mortal
LAW All men are mortal

ABDUCTION
RESULT Socrates is dead but: RESULT Socrates is dead
LAW All men are mortal LAW All fruit flies are mortal
CASE Socrates is a man CASE Socrates is a fruit fly

Children take what they hear around them (‘results’), apply innate knowledge
(‘laws’) and construe a grammar (the ‘case’), then they test the grammar by
induction (checking it against more data) and deduction (speaking and
succeeding or not).

2.4 Pragmatic explanations

Functionalist view:

• change is the result of ‘choice-making on the part of speakers / writers in

interactional negotiation with addressees/readers’ (Traugott 2002: 21)
• language change as the result of use (by adults not children)
• inferences become fixed and grammaticalised

2.4.1 English go
English Stage 1 I am going to fetch water from the river
implies ‘in the future I will get water from the river’

*I am going to get better is not possible.

English Stage 2 going to has inherent (not inferred) future meaning, so

I am going to fetch water from the river
means ‘in the future I will get water from the river’

and I am going to get better

means ‘in the future I will get better’

3 REANALYSIS
Reanalysis = ‘a mechanism which changes the underlying structure of a syntactic
pattern and which does not involve any immediate or intrinsic modification of its

3 DWW
surface manifestation’ (Harris & Campbell 1995: 61, definition goes back to
Langacker 1977).

3.1 English for preposition > complementiser

(7) [It is better for me] [to slay myself than to be violated thus] Middle
English
(8) [It is better] [[for me to slay myself] than to be violated thus] Modern
English

The constituency of Modern English can be shown by:

(9) [For me to slay myself] would be better than to be violated thus.

3.2 Changes in grammatical relations: English think

Middle English: Him thinks The king thinks Me thinks

Modern English He thinks The king thinks I think

• for changes category (preposition to complementiser)

• the clause boundary and constituency relations shift

3.3 Timberlake’s (1977) study of reanalysis and actualisation: Finnish

participial clauses

The claims:
• there are two aspects to syntactic change

reanalysis (new underlying relationships)

actualisation (gradual mapping out of the consequences of the reanalysis)

≅ extension ‘a change in the surface manifestation of a syntactic pattern

that does not involve immediate or intrinsic modification of underlying
structure’

• actualisation of change moves from less to more marked contexts

The grammar determines the output. Learners formulate a grammar that

produces the given output. They do not have to formulate a grammar identical to
that of others.

Sometimes a given output could be produced by different grammars.

Change/reanalysis is a shift between two such grammars. Actualisation is
subsequent concrete changes in ‘norms’.

Finnish participial clauses

Finnish objects may appear in a number of cases, broadly:

• partitive if the clause is negative

• accusative if the verb has an overt subject
• nominative if the verb has no overt subject

Verbs of cognition (‘think’, ‘notice’) are and were followed by a participial clause.
In older Finnish the subject of the participle went into the same cases as objects

4 DWW
[partitive with certain verbs/for incomplete actions etc.; nominative if the verb
has no overt subject; otherwise accusative]:

(10) eike lwle site syndi oleuan

not think this.PART sin being
‘Does one not think this to be a sin?’

(11) seurakunnan hen lupasi pysyueisen oleuan

congregation.ACC he promised long-lasting being
‘He promised the congregation to be (i.e. would be) long-lasting’

(12) …homaitan se tauara ia Jumalan Lahia poiseleua

observed goods.NOM and God’s gift.NOM lacking
‘It is observed the goods and the gift of God to be (i.e. are) lacking.’

This suggests:
• the underlying subject of the subordinate clause (‘this’, ‘the congregation’ and
‘the goods and the gift of God’) raised to become the object of the main clause
(subject-to-object raising)
• therefore it took the case appropriate to the object of the main clause

In Modern Finnish, the subject of such clauses is always genitive, which suggests
that it’s not an object:

(13) En sanonut lapsen tulevan Older Finnish would be:

not say child.GEN coming ____
‘I did not say the child would come.’

(14) Näin lapsen panevan kirjeen taskuunsa

saw child.GEN putting letter in-his-pocket ____
‘I saw the child putting the letter in his pocket.’

(15) Lapsen huomattiin varastavan parhaan hevosen

child.GEN observed stealing best horse ____
‘The child was observed stealing the best horse.’

How has this reanalysis come about?

The accusative ending -m merged with the genitive ending -n in singular nouns
but not in plural/pronouns. This makes one group of examples ambiguous [active
verb with singular noun phrase subject]. So a grammar with underlying genitives
and individual rules for exceptions in each position would work.

Grammar 1 Subject V Objecti [ti V]

then apply general rules for assigning case to objects

Grammar 2 Subject V [Subject V]

then apply general rule that subject of a participle is genitive
plus exception rules for the special cases

Subsequently the exception rules are lost (subsequent generations fail to learn
them).

5 DWW
3.4 Features of and issues in reanalysis
3.4.1 Reanalysis needs ‘ambiguity’
Reanalysis needs ambiguity in the surface output, but how much is not clear. In
both the last two examples, there is ambiguity in only one environment. Perhaps
ambiguity in the least marked (and statistically most frequent) environment
helps.

It is still surprising that the ambiguity inherent in only one variant of

the construction could be taken as representative of the whole
construction. (Timberlake 1977: 149).

A subset of the tokens of a particular constructional type must be

open to the possibility of multiple structural analyses, where one
potential analysis is the old one (applicable to all tokens) and the
other potential analysis is the new one (applicable to a subset).
(Harris & Campbell 1995: 72)

3.4.2 Reanalysis may be fed

Reanalysis may be facilitated by change in other parts of the grammar:
• in the Finnish example, phonological change leads to the merger (syncretism)
of accusative and genitive for some nouns, providing the ambiguity needed for
reanalysis.

Compare also the appearance of French question marker ti, which may have been
facilitated by the phonological loss of the l in -t-il in:

(16) Aime-t-il?
love-T-he
(17) Tu vas ti?
you go-2S Q

3.4.3 Reanalysis and directionality: The chicken and egg problem

(change is the chicken, reanalysis is the egg)
Two views:
1. Reanalysis occurs late, only after a series of unrelated changes which
complicate the grammar. If so, why should the complications arise in the first
place? How did the change maintain its directionality, or, in the terms of
Andersen (1990), what is the structure of drift?

2. Actualisation follows reanalysis: directionality maintained through exception /

adaptive rules (see Andersen 1973, Disterheft 1990).

4 MORE GENERAL ISSUE: DISCRETENESS AND GRADUALNESS

Syntactic change is gradual in that:

(i) the grammar gradually adjusts to the new system through a series of
actualisations in different syntactic environment (Timberlake’s 1977 example);

(ii) some syntactic phenomena spread from lexical item to lexical item (Naro &
Lemle 1976; Warner 1982: 117–23; Fischer & Van der Leek 1987);

6 DWW
(iii) extensions may remove one exception at a time (cf. Harris & Campbell’s
constraint that extension of a rule is limited to removing conditions on its
operation);

(iv) the frequency of a new phenomenon may increase gradually over time (Kroch
1989).

Syntactic change is discrete in that:

(i) the grammar is constructed anew by each generation of speakers;

(ii) the process of reanalysis is itself discrete: any individual speaker has either
the new analysis or the old one.

FURTHER READING
The historical linguistics textbooks vary enormously in quality on syntactic
change. McMahon and (especially) Campbell are the best, but you will need to
look beyond them.

* Harris, Alice C., and Lyle Campbell. 1995. Historical syntax in cross-linguistic
perspective. Cambridge: Cambridge University Press, esp. chapters 3–5.
* Timberlake, Alan. 1977. Reanalysis and actualization in syntactic change. In
Mechanisms of syntactic change, edited by C. N. Li. Austin: University of
Texas Press, 141–77.
* Hopper, Paul J., & Elizabeth Closs Traugott. 2003. Grammaticalization.
Cambridge: Cambridge University Press [Use the 2003 edition if possible.
For reanalysis, look at chapter 3.]

For further reading (and full references), see the supervision assignment on
mechanisms of syntactic change on the website.

The English stress alternation example is from:

Sherman, Donald. 1975. Noun-verb stress alternation: An example of the lexical
diffusion of sound change in English. Linguistics 159: 43–71.

The development of English consonants is from:

Lucas, Peter J. 1991. Some aspects of the historical development of English
consonant phonemes. Transactions of the Philological Society 89: 37–64.

7 DWW