Language from other languages or dialects, for


change may occur under external influence from other languages or dialects, for
example, by the process of language contact, or they can change through
internal factors, the main focus of this essay. I will discuss the work of
Lightfoot (1991, 2007) and Chomsky (1965, 80) looking at syntactic change as a
result of child language acquisition from a generative perspective. In order to
understand the relationship between language change and grammar, I will investigate
how children acquire their native language in the first place. Focusing on the
history of English, I will demonstrate diachrony in syntax using examples from
Old English (OE). Furthermore, I will contrast generative approaches with the
view that all language change is a result of contact.

approaches to language acquisition suggest that the syntactic knowledge that
children acquire is underdetermined by input alone, rather learners can access
an innate language acquisition device (Chomsky 1965). Chomsky (1980) defined an
important argument for the generative perspective known as the poverty of the
stimulus (POS) argument which suggests that the input that learners are exposed
to isn’t rich enough for them to acquire every feature of the target grammar
yet they are still able to do so. According to McMahon (1994), the grammar of
each generation is not passed on from adult speakers to their language-learning
children but that each speakers’ grammar is self-contained and potentially
different. The POS suggests that this new target grammatical structure must be
acquired due to some internal linguistic property of the learner. Chomsky
(1965) discusses the idea of a Universal Grammar (UG), a set of structural
rules that are innate to humans and analysable in all languages. The UG
interacts with external Primary Linguistic Data (PLD) creating a personal,
internalised mental grammar, known as competence (McMahon 1994).

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!

order now

regards to syntactic change, Andersen (1973), an earlier generative grammarian,
also proposed that children exposed to the linguistic data of their environment
may produce a grammar that is minimally different to that of the input. This
may only be a slight difference, unrecognisable by the learner making them less
likely to repair it resulting in subtle language change over time. According to
Kroch (2001) language change can be defined as a failure in the transmission
across time of linguistic features. In the case of syntactic features, grammars
change when there is enough change in the learners’ output to set new
grammatical parameters and if there is not then the grammars are stably
transmitted (Lightfoot 1991). Such failures of transmission however can be seen
as failures of learning in child language acquisition since the grammar that
the learner fails to acquire has been learnable in the immediate past. However
these failures do occur leading to language change perhaps due to a change in
the character of the evidence that the child is exposed to or in some
difference in the learner (Kroch 2001).

generative linguistics, syntactic changes in language are assumed to occur
between two generations. The idea is that a new generation acquires some
components of a parents’ output. It is therefore important to explore how
children acquire language in order to understand how language change occurs. In
other words, understanding more about learners and their interaction with the
structural properties of a language allow for a better understanding of
diachronic syntax. According to Lightfoot and Westergaard (2007), learners are
exposed to PLD which refers to simple structures expressed in external language
(E-language) and in response they are able to grow their own internal language.
The structures heard by the learners are known as cues, found in the PLD which
can be analysed, given what the learner already knows, only if a particular cue
is present. In other words, the cue-based approach contains pieces of syntactic
structure which learner’s form in their internalised grammar (I-language) when
they hear certain triggers in the E-language. It differentiates between the
abundance of unanalysed data that learners hear and the mental grammar that
grows in the mind of the learner. According to this approach, we can assume
that there are large-scale word order parameters provided by UG that are
formulated as cues, for example, the head parameter such as VO vs. OV. The cue
for OV syntax is VPDP V indicating a verb-final word order. The
cue itself is not found in the input but is expressed by particular input
strings encouraging a syntactic analysis in the child grammar (Westergaard
2009). This cue-based approach allows us to view the essence of a historical
change as the point when cues are expressed differently in the E-language
provoking the growth of new I-languages in learners.

At the
level of syntax, language change often occurs within the history of a single
language. I will discuss the shift from Object-Verb (OV) to Verb-Object (VO)
order in Old English in order to illustrate syntactic change as a result of
child language acquisition. Lightfoot (1991) claims that this shift is due to a
parametric change. He defines child language learners as of degree-0
learnability and suggests that only simple structures are involved in
acquisition. He argued that degree-0 learners set parameters on the basis of
main clause evidence as they are only sensitive to root clauses and unembedded
binding domains (Kroch 2001). Language change, namely the shift from OV to VO
order in OE is an appropriate testing ground for such a restrictive hypothesis.
With regard to language change over time, the population of Old English
speakers shifts from an OV-biased distribution around 1000 A.D., as shown in
the examples below, to a VO-biased distribution around 1200 A.D. (PPCME2
Corpus, Kroch & Taylor, 2000).

(1a) heSubj
GodeObj þancodeTensedVerb

God       thanked

        ‘He thanked God’.

        (Beowulf, 625, ~1100 A.D.)

(1b)  & mid his stefnePP heSubj
awecUTensedVerb deadeObj to lifePP

         & with his stem        he     
awakened          the-dead to life

         ‘And with his stem, he awakened the
dead to life’.

        (James the Greater, 30.31, ~1150 A.D.)

(Pearl and
Weinberg 2007)

(1991) proposes that this type of change is a result of a misalignment of the
child’s hypothesis and the adult’s analysis of the same data. He claims that
children have access to triggers in frequently occurring linguistic data which
allows them to set appropriate parameters, some of which are evident in main
clauses of OE. Over time, however, OV order in main clauses declined in
frequency until they were no longer frequent enough for learners to recognise,
making them “unlearnable” (McMahon 1994). Instead, reanalysis took place resetting
the parameter in both main and embedded clauses, however it is important to
note that if learners had access to subordinate clauses as input data,
reanalysis wouldn’t have happened at all and many OV structures would have
still existed. With the view that children are indeed degree-0 learners then we
can conclude that while the clues to OV order were becoming less clear and
frequent, VO order was increasing in frequency, eventually causing a parametric
change. While Lightfoot argues that this change was due to a catastrophic
decline in the frequency of OV order in subordinate clauses, Pintzuk (1991)
demonstrates how the discontinuity may be accounted for as a sociolinguistic
phenomenon rather than a grammatical one. The Constant Rate Effect refers to
the rate of replacement remaining the same in all contexts affected by the
change when language change is a result of grammar competition (Kroch 1989). This
shift is an example of this hypothesis as both main and embedded clauses
experience a gradual rise of VO order with the same rate of increase. VO order
in OE can therefore be described as an independent parametric option was not a
transformational variant of underlying OV order like in modern Dutch and German
(Pintzuk 1991).

I will
also discuss a structural shift which served to make Early Modern English (EME)
differ from Middle English (ME). That is, the introduction of a new
inflectional category which allowed for modal verbs as we know them today such
as might, must, shall and should. The following examples from Lightfoot (2007) demonstrate
the new behaviour of modal auxiliaries in EME:

(2a) He
will try to understand.

(2b) ?He will can understand.

illustrates how lexical verbs may be used with a modal but not with a second
modal as in (2b). EME demonstrates the use of combinations of modal auxiliaries
as in (3).

(3) “
letters shall may come unto your
grace’s hands” (1532, Cranmer, Letters)

auxiliaries have always been a distinctive class of verbs with certain
characteristics, for example, they have never had the –s ending, a defining property
of English verbal morphology (Lightfoot 2007). In the OE period, however, they
had a much wider range of verbal inflections and more characteristics than
typical lexical verbs such as having objects and tensed clauses complements.
Over time, they acquired too many exceptional characteristics to be learnable
as lexical verbs and as a result, they have undergone reanalysis and became
members of this new category due to a simplification of morphology in ME.
Lightfoot (1979) suggests that this reanalysis is due to the Transparency
Principle (TP), a universal principle responsible for syntactic change. The
principle can be defined as a control on abstractness in syntax restricting the
amount of obscurity or exceptionality there is in a grammar. In other words,
the TP requires that grammar must be minimally complex and when it is too
obscure or far from the surface structure then reanalysis will occur as it has
in the above example with modal auxiliaries.

I have
demonstrated how certain generative approaches claim that the essence of
syntactic change is to be found in the process of child language acquisition
however language change may also occur under external influence too considering
the vast amount of speakers who are at least bilingual. To contrast, I will
discuss the process of language contact, defined as the transfer of elements
from one language to another, a process which relies on bilingualism (McMahon
1994). The view that all language change is a result of language contact argues
in favour of the constructivist theory that children acquire language from
input only in contrast to the self-contained, internally motivated influence
discussed in this essay. This view supports the idea that early child language
production is not rule governed but rather the input that learners have been
frequently exposed to allows for imitation of simple strings of elements. Contact-induced
language change is due to imperfect learning suggesting that adult L2 learners
expose children to an imperfect version of the second language who are now
native speakers of this “foreign dialect” (Kroch 2001). In comparison to
inaccuracies in child language acquisition, this is a much clearer example of
the cause of imperfect transmission. However with regards to syntactic change,
grammatical features are not often borrowed and constraints exist against this
kind of change (Winford 2003) supporting the view that not all language change
is contact-induced.

From the
perspective of generative linguistics, I have illustrated how English has
evolved over time with regards to syntactic change. I have demonstrated how
children acquire their grammar with influences from external language
triggering new internal languages taking into account the different
environments, experiences and new PLD different learners are exposed to
indicating the possibility of new grammars emerging. I have also mentioned only
a couple of the many changes in the history of English to illustrate the role
that child language acquisition has played in syntactic language change.
Lastly, it is important to note that many processes of language change exist
but it is equally important not to rule out internal influences, especially
with regards to syntactic change.