Language

This module on the nature of language aims to provide the novice and
even experienced teacher with a broad and accessible picture of language
as a formal system. As such, it covers topics such as the nature of words,
sounds, and syntax. The module places particular emphasis on the
abstract and complex nature of language and how it does not resemble
typical pedagogical rules and so-called rules of thumb often used with
language learners.

Bill VanPatten is professor of Spanish and second language studies at

Michigan State University.

[Link] i 8/15/2015 6:00:45 PM

 The Routledge E-Modules on Contemporary
Language Teaching
Series Editors: Bill VanPatten, Michigan State
University, USA, and Gregory D. Keating,
San Diego State University, USA

The Routledge E-Modules on Contemporary Language Teaching series is

an exciting and innovative approach to topics for the novice or in-training
teacher of languages. Written in an easily accessible style and delivered in
e-format only, specialists and experts provide the latest thinking on a variety
of issues that form the foundation of language-teacher knowledge and prac-
tice: the nature of language and communication, second language acquisi-
tion, interactive tasks, assessment, focus on form, vocabulary development,
and technology in language teaching, among many others. Each module
serves as a self-contained unit to be used on its own or as part of an intro-
ductory course on language teaching. Instructors may “mix and match”
modules to create their own readings for a course on language teaching.
The modules may serve as primary reading or as supplemental reading,
with each module offering points of reflection, discussion questions, self-
quizzes, and a reading list for those who wish to delve further into the topic.

Language
Bill VanPatten

Communication and Skill

Bill VanPatten

Second Language Acquisition

The Basics
Gregory D. Keating

Vocabulary in Language Teaching

Joe Barcroft

Interactive Tasks
Michael J. Leeser and Justin P. White

Focusing on Form in Language Instruction

Wynne Wong and Daphnée Simard

Technology in Language Learning

An Overview
Bryan Smith

[Link] ii 8/15/2015 6:00:46 PM

 Language

Bill VanPatten

[Link] iii 8/15/2015 6:00:46 PM

 First published 2016
by Routledge
711 Third Avenue, New York, NY 10017
and by Routledge
2 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN
Routledge is an imprint of the Taylor & Francis Group,
an informa business
© 2016 Taylor & Francis
The right of Bill VanPatten to be identified as author of this work has
been asserted by him in accordance with sections 77 and 78 of the
Copyright, Designs and Patents Act 1988.
All rights reserved. No part of this book may be reprinted or
reproduced or utilized in any form or by any electronic, mechanical,
or other means, now known or hereafter invented, including
photocopying and recording, or in any information storage or
retrieval system, without permission in writing from the publishers.
Trademark notice: Product or corporate names may be trademarks
or registered trademarks, and are used only for identification and
explanation without intent to infringe.
ISBN: 978-1-315-67959-4 (ebk)

Typeset in Sabon
by Apex CoVantage, LLC

[Link] iv 8/15/2015 6:00:46 PM

 Language
Bill VanPatten
MICHIGAN STATE UNIVERSITY

Overview
In this module you will explore the following topics:

• language as mental representation

• the abstract and implicit nature of mental representation
• the components of language (e.g., lexicon, phonology, mor-
phology, syntax, pragmatics)
• how language as mental representation is distinct from peda-
gogical grammar and prescriptive grammar
• how the above information can be used to consider the nature
of second language acquisition

Language is a uniquely human trait. By this we mean that of all the liv-
ing species on Earth, only humans possess a faculty for language and
only humans communicate with language—although we also know that
humans communicate with things other than language (as do other spe-
cies). Although we might share with bees that we can signal meaning
through a dance move, and although we might share with some birds the
use of whistle noises to convey an idea, no animal other than humans
possesses the ability to modify a noun with a relative clause or the ability
to determine that John seems sad is a fine sentence but John seems asleep
is not.
So, why is language a uniquely human trait? What is language? In
this module we will explore language as a formal and at times very
abstract system. We will go beyond the layman’s concept of language
expressed in terms such as “don’t use bad language” and “I speak
three languages.” We will go beyond the common conceptualization
of language as a set of rules such as “Use whom instead of who after
prepositions.” We will go beyond the widely held belief that language is

[Link] 1 8/15/2015 6:00:46 PM

 2 Bill VanPatten
communication—although readers can explore the nature of communi-
cation in another module.

Language as Mental Representation

Every speaker of a language is also a “knower” of language. That is,
underlying every act of language use there lurks implicit knowledge
about language. This knowledge is not the everyday kind of knowl-
edge such as “Don’t say ain’t” and “Verbs must agree with their sub-
jects (in some languages).” Knowledge of language can involve complex
and abstract features, as we will see later. Important here is that this
underlying knowledge is implicit. Let’s call this knowledge “mental rep-
resentation.” What does implicit mean? Let’s begin with a non-linguistic
example. Every reader of this module knows what “round” is. Marbles
are round. Baseballs are round. Doughnuts are (generally) round. U.S.
coins are round. Wedding bands are round. But what is the definition of
round? As a reader tries to answer this question, he or she will sit and
think, attempting to come up with an answer. But this sitting and think-
ing suggests that the actual knowledge underlying “round” is deep and
implicit, probably even abstract. The person trying to come up with a
definition has to reason it out. I’ve tried this question on audiences when
I give lectures and inevitably all responses to the question are wrong.
Thus, even if we can make our knowledge somehow explicit through
reasoned thought, we are often wrong for concepts such as “round.” (For
the reader dying of curiosity, “round” means that each surface point is
equidistant from a center point.)
When it comes to language, the term implicit suggests that men-
tal representation consists of knowledge that a person cannot easily
describe or cannot describe at all. Although we will explore this in
detail in the sections that follow, here is a good example. In English,
we can say You told Harry about which book? as well as Which book
did you tell Harry about? That is, the phrase which book can move
from one spot to another spot. But you can only say You told Harry
about which politician caught in a sex scandal? and not Which politi-
cian did you tell Harry about caught in a sex scandal? That is, which
politician is not free to move the way which book was in the previous
example. The question is, “Why?” Every speaker of English would rec-
ognize the Which politician . . .? sentence as bad or weird, but when
asked why, that person would fumble for an answer or simply say, “I
don’t know. It just sounds bad.” When someone says, “I don’t know.
It just sounds bad,” that person is indicating that he or she is relying
on implicit knowledge; that is, his or her mental representation. The
knowledge is clearly there; the person is simply unable to articulate it.
Again, we will see more examples of this later and will drive the point

[Link] 2 8/15/2015 6:00:46 PM

 Language 3
home when we talk about pedagogical and prescriptive grammars much
later in this module.
As stated above, language is also complex and abstract. It is abstract
because there is much about the nature of language (our implicit knowl-
edge) that defies simple and easy explanations. Before looking at a lin-
guistic example, let’s return to the definition of “round”: each surface
point is equidistant from a center point. The reader might think, “Of
course, that’s so simple.” But it really isn’t that simple when we begin
to deconstruct the definition. For example, what’s a “surface”? What’s
a “point”? What’s a “center”? What does “equidistant” mean? And so
on. Each of the constructs within the definition of “round” also involves
implicit, abstract knowledge, making the overall definition of “round”
much more complex than what meets the eye. The same is true for lan-
guage. Again, we will see examples of this in the sections that follow, but
for the present purposes, we can illustrate linguistic abstractness in the
following example. English allows contractions such as I’ve done it and
Who do you wanna invite to the party? But English does not allow the
same contractions in Should I’ve done it? and Who do you wanna fire
John from his job? Contractions are prohibited in English under certain
conditions, namely, when an element has moved from one part of the sen-
tence to another, leaving what we call a trace behind. For example, I’ve
done it is a contraction of I have done it. Nothing intervenes between I
and have, so contraction is possible. But Should I have done it? is derived
from I should have done it. When the modal should moves, it leaves
behind a hidden (i.e., invisible and inaudible) copy of itself known as a
trace, which linguists represent as a small t. Here is an example. Should
I t have done it? This little t, which marks the spot left by should, blocks
the contraction of I and have. Here is what this looks like if linguists were
to draw a tree to represent the sentence (ignore all the symbols for now
and just focus on how the movement of should is depicted).

[Link] 3 8/15/2015 6:00:46 PM

 4 Bill VanPatten
The idea of moving items, leaving traces, and so on, involves a level
of abstraction we don’t normally think about every day when it comes
to language. At the same time, the blocking of contractions is complex:
it involves multiple aspects of language (e.g., movement, restrictions on
movement, traces and trace theory).
A major issue in second language research has been to what extent
learners develop an underlying implicit and abstract/complex system.
The answer is that they do, and parts of the system may be indistinguish-
able from what native speakers have in their representations while other
parts are different. Another module on second language acquisition will
touch upon this idea in more detail and will indicate further readings
for you. In this module, we will merely point out now and then what we
know about what second language learners have in their heads.
So far, we have seen complexity only at the level of how a sentence
operates, but language overall is complex because it has so many differ-
ent components that work together to form what we call sentences or
utterances. To illustrate this complexity, we will briefly talk about some
major components of language: phonology, lexicon, morphology, syntax,
and pragmatics, to name some. We begin with phonology.

Phonology
In English, we pluralize nouns by adding –s; at least this is what we say.
But try saying some words out loud: cat/cats, dog/dogs, house/houses.
If you say the words and listen, you will notice that the first pluralized
word (‘cats’) ends in what we might call a typical ‘s’ sound. However, the
next pluralized word (‘dogs’) ends in something that sounds like the first
sound of zebra. Finally, the third pluralized word ends in neither sound
and actually makes its own syllable, sounding something like fizz with the
‘f’ removed. Yet an English speaker is never confused and always knows
when a word is plural or not and what the sound should be that makes a
noun plural. This is because the English speaker possesses a mental repre-
sentation of what a plural ending is, along with what happens depending
on what other sounds that ending comes into contact with. The mental
representation of sounds is called the phonological representation while
the actual sound produced is called its phonetic realization. Thus, the
phonological representation for plural on nouns might be something like
/-s/, but it has three phonetic realizations: [-s], [-z], [-ɪz].
Let’s examine something else, this time from Spanish. Spanish pos-
sesses both an ‘m’ sound and an ‘n’ sound (similar to English) so that
como (‘like,’ ‘as,’ ‘how’) and cono (‘cone’) are two different words with
different meanings. We call sounds that result in different meanings pho-
nemes; for example, ‘t’ in tie and ‘d’ in die are phonemes because they
result in meaning differences when all other sounds in the same word are
the same, even though ‘d’ and ‘t’ in and of themselves don’t carry any

[Link] 4 8/15/2015 6:00:46 PM

 Language 5
meaning. In Spanish, however, the phonemes represented by m and n are
neutralized in certain positions, normally before certain consonants. By
neutralized, we mean that the substitution of one for the other makes
no difference in the word. When an ‘n’ is immediately followed by a ‘b’
with no pause, for example, it sounds like an ‘m’. Thus, the sequence con
botas (‘with boots’) actually sounds like ‘combotas’ when spoken. If we
use slash marks to indicate phonological representation and brackets to
indicate phonetic representation as we did for the plural markers in the
paragraph above, what occurs in this Spanish example is that /konbotas/
is realized as [kombotas]. (When in spoken language, one sound takes
on the characteristics of an adjacent sound, this is called assimilation.
If you make a ‘b’ sound you will see that your lips come together with
a slight rounding. Now make an ‘n’ sound. You will see your lips are
in a different position. But if you make an ‘m’ sound you will see that
your lips have a similar position to the ‘b’ sound. This is what we mean
by assimilation and one sound taking on the characteristics of another
sound.) Languages vary as to what the immediate linguistic environment
does to a phoneme. Japanese allows devoicing (lack of vocal chord vibra-
tion) in the production of high vowels (‘ee’ and ‘oo’ sounds, as in English
‘leak’ and ‘cool’) in certain contexts where they are adjacent to voice-
less consonants (e.g., /s/ and /k/, in which there is no vocal chord vibra-
tion). Turkish has vowel “harmony,” in which vowels added because of
morphological inflections (see later section in this module) take on the
characteristics of the vowels in the root word. In short, sounds may be
“malleable” depending on the other sounds they come into contact with.
So, in any given language, there is an inventory of phonemes, there is
a phonological representation that a language’s speakers carry around in
their heads, and there is a phonetic representation that governs how those
sounds are actually spoken. Phonologists study these particular aspects
of language. They also study the nature of syllables. Languages vary in
terms of what they allow as sound sequences when forming syllables.
English is a language that allows both open and closed syllables. Open
syllables end in a vowel as in the simple word ‘no’ /no/ and the first syl-
lable of the word ‘chicken’ /tʃɪ-kn/. Closed syllables end in a consonant
as in the simple word ‘ten’ /tn/ and in the second syllable of ‘chicken’
/tʃɪ-kn/. Spanish and French also allow both open and closed syllables,
but they differ from English in that there is a marked preference in these
two languages for open syllables such that when spoken, the sounds of
some words are carried over to the next word in order to create open
syllables. In Spanish, the sequence con Ana (‘with Ana’) is most often
articulated such that the syllables are [ko-na-na]. That is, the [n] of [kon]
migrates from its place as the end of a syllable to be the beginning of the
next syllable.
At the same time that Spanish and French allow closed syllables, they
differ from English in the number of consonants they allow to close a

[Link] 5 8/15/2015 6:00:46 PM

 6 Bill VanPatten
syllable. In both languages the maximum number of consonants is one.
This is why in Spanish plurals, whenever the word ends in a consonant,
the plural sound is always [-es] as in señor and señores (cf., salsa and
salsas). How many consonants can clump together to end a syllable in
English? Pronounce the verb glimpsed. You will hear the syllable close
with this phonetic sequence: [-mpst]. This difference in the closure of syl-
lables partially explains why speakers of some languages have difficulty
with English noun and verb inflections. Imagine the Spanish or Italian
speaker in the initial stages of learning English and trying to add a [z] to
the end of dog so that the syllable closes with [-gz], an impossible syllable-
final structure in their languages. Other languages don’t allow consonant
clusters to begin a syllable, as in the case of Chinese. All Chinese syllables
must begin with a simple consonant or a vowel. In short, when people
learn other languages, they have to learn the inventory of phonemes and
they may also have to develop a new mental representation for phonologi-
cal sequences that don’t exist in their own language. They may also have
to learn how these sounds are realized in real time speech.

Reflection
Italian does not have the same inventory of vowels that English
does. There is no vowel sound that corresponds to the vowel in
“cook” and “book.” The closest vowel sound in Italian to this
vowel is the one found in “kook” and “spook.” English speak-
ers easily discriminate and pronounce “cook” and “kook,” for
example. Imagine an Italian speaker who has lived in the U.S. for
some time. In speaking, she pronounces “cook” and “kook” the
same, with the pronunciation of both being “kook.” You give her
a discrimination test such as “Listen to what I say and then tell me
which sentence goes with what you hear. ‘She’s a kook.’ A. She’s
good in the kitchen. B. She’s crazy.” This Italian speaker says “B,”
which is the correct answer. What do you make of this in terms of
the concepts of phonological representation and phonetic realiza-
tion for this second language learner?

To conclude, phonology is not just about sounds but also about how
sounds behave. Phonology is as much about the combinatorial con-
straints on sounds as it is about what the sounds themselves are. Note
that we have not touched upon the issues of stress, pitch, and tonal pat-
terns in languages—all of which form a part of the phonological repre-
sentation that people have for a language. In and of itself, the phonology
of language is complex, and when it comes to sounds, much more goes
on in people’s heads than what they realize.

[Link] 6 8/15/2015 6:00:47 PM

 Language 7
Lexicon
Every speaker of a language possesses what we call a mental lexicon, a
sort of dictionary, although quite unlike the physical and online diction-
aries that we use to check meanings and spellings. This mental lexicon
consists of words and phrases, connections between them, and all the
properties associated with these words and phrases. When most people
think about words, they think about meanings, but words consist of
much more than just their meanings. Let’s take the word money. As a
lexical entry in a person’s head, this word has a phonological representa-
tion (i.e., a set of phonemes sequenced in a certain way, as in /mʌni/). Lin-
guists often refer to this phonological representation as a lemma, which
simply means the form of the word. This lemma is accessed from the
mental lexicon when a person either wants to use the word money or
hears it spoken by someone else.
But this word also has grammatical properties, namely what kind of
word it is (noun, verb, adjective, and so on), which determines where it
can be slotted into a sentence, and information about how it can combine
with other words. In the case of money, it is classified as a noun but it is
also classified as a mass noun, and thus is constrained by what kinds of
determiners (articles) can precede it. As a mass noun, it can form part of a
phrase with some, any, and the, but it cannot form part of a phrase with
a; for example, Gimme some money, Do you have any money?, Show
me the money, but *I don’t have a money. (In linguistics, an asterisk is
used to indicate a sentence, phrase, or form of a word is not permissible
in a given language.) So, the word money in a speaker’s mental lexicon
is indexed with features such as <+N> <-V> <+mass>1 along with the
phonological properties /mʌni/.
Grammatical properties of words become extremely important with
verbs. Encoded into the mental lexicon with a verb are its syntactic prop-
erties. A speaker of English knows, for example, that Put this jacket on
your little brother is a fine sentence, as is Put it on your little brother. But
that speaker also knows that *Put this jacket to your little brother is not
fine (although it is in other languages) and that *Put this jacket your little
brother is also not allowed in English. A speaker of English also knows
that Put it on is fine but *Put it is not. In the mental representation
English speakers have for put is the fact that it normally requires three
arguments: the person or thing who puts (called the ‘agent’), the thing
that is put (called the ‘theme’), and where the thing is put (called the
‘goal’). In the final case, this argument is realized as a prepositional
phrase, the absence of which makes the sentence ungrammatical. Note,
however, that “equivalent words” in other languages may not end up
having the same grammatical properties of English even if the verb has
the same arguments. A case in point: Japanese oku, which is equivalent
to “put.” This Japanese verb also has three semantic arguments, but in
Japanese syntax, the expression of the theme (object) in the syntax is

[Link] 7 8/15/2015 6:00:47 PM

 8 Bill VanPatten
not required under certain conditions; in English some expression of the
theme is always required. All of this information is stored with each lexi-
cal item so that when the lexical item is used to form sentences, “stuff can
match” during sentence generation.
Arguments are tied to the semantic properties of words, semantics
referring to the meanings of words. A part of the meaning of put is some-
thing like “X causes Y to move to location Z.” Thus, the underlying
semantics of a verb determine its argument structure, which in turn has
consequences for the way the verb must be used in a sentence. Because
of its semantics, ‘put’ has three entities normally involved in the event:
the putter, the thing put, and the place where the thing put winds up. As
a quite different example, let’s look at the verb seem. In the introduction
to this module we saw that John seems sad is a fine sentence but *John
seems asleep is not (although John seems to be asleep is fine). Also ruled
out are sentences like *John seems a professor and *John seems the food.
Within the semantics of seem is the notion of experience, that is, seem
is not an action word like put. Instead, its semantic representation con-
tains something like “X experiences Y, where Y is a condition subject to
change.” As such, seems takes only one argument (in this case called ‘the
experiencer’), which is generally realized as the subject of the sentence.
Because it cannot take any other argument, *John seems the food is disal-
lowed, as is *John seems a professor. A similar example is the verb die.
The subject of die is not an agent who brings about the event of dying,
but something that experiences death. Thus, Her mother is dying of a
rare disease is fine but A rare disease is dying her mother is not.
Some lexical elements of language lack semantic properties; that is,
they are devoid of meaning and serve grammatical functions only. One
very good example is the auxiliary do in English—used to make yes/
no questions (Do you want to go?), negative sentences (I don’t know),
emphatics (Ah, but I DO know!), and other structures. There is nothing
out there in the real world that we can point to that is a do or that is
doing. But do exists in the English speaker’s mental lexicon and it exists
with important information, largely grammatical. Its sole function is to
carry tense features (past or present as in does versus did). We will see
how this works in more detail later in the section on syntax.
Certain words that begin clauses are also meaningless, such as the that
in I think that linguistics is interesting. Its sole function is to introduce
a clause, and as such, under certain conditions it is not even necessary:
I think linguistics is interesting. Such meaningless words are stored in
the lexicon like any other word, but their inventories do not contain
anything about meaning, only their underlying grammatical properties.
In short, the mental lexicon we carry in our heads contains all words,
whether the words have semantic information or not. And all of those
words have grammatical information as part of their lexical entry.
Connections exist in the mental lexicon such that related words are
linked (e.g., brother/sister, sun/moon/sky, push/shove, pull/yank, think/

[Link] 8 8/15/2015 6:00:47 PM

 Language 9
believe/know/cogitate/thought/idea/notion/concept). These links are based
on the semantics (meanings) of words, but the fact that words are linked
semantically does not mean they share the same grammatical properties.
Both John thinks many things and John believes many things are allowed,
whereas John believes himself to be a good man is fine but *John thinks
himself to be a good man is not. Thus, words have both meaning and
grammatical properties resulting in requirements that must be fulfilled in
sentence structure.
Earlier we said that the mental lexicon consists of words and phrases.
So far, we’ve only talked about words. But everyday language use con-
tains a lot of what is called “formulaic language,” phrases and chunks
of language that are routinized and used in restricted ways. Examples
include How’s it going?, What’s up?, Hell if I know, and Beats me. These
are true phrases in the lexicon as they are not stored as a set of words
but as indivisible units such as /howzɪtgoɪn/ and /bitsmi/. Storing such
routinized chunks of language in this way is what helps make everyday
conversation fluid and fast.

Reflection
Think about the following sentences: (1) John broke the vase;
(2) The vase broke; (3) The vase fell; (4) *John falled/fell the vase.
What are the differences between break and fall in terms of their
underlying semantics and argument structure? Why do you think
(2) is possible but (4) is not?
Also, think about your own learning of another language. When
learning words, did you focus on their meanings or did you focus also
on all of their underlying grammatical properties as outlined here?

Morphology
Let’s begin this section with a quick test. We know that re- can be com-
bined with verbs to mean something akin to “do again,” as in re + paint =
repaint “to paint again.”

(1) John painted the walls but had second thoughts about the color. So
he repainted them the next week.

Here’s the test: do you like all of the following sentences equally well?
If not, which ones don’t you like or which ones sound better than others?

(2) Roberta recovered the sofa and now it looks great.

(3) The patient slept all night but was so tired he reslept the next day.
(4) The child petted the dog and when it wagged its tail, he repetted it.

[Link] 9 8/15/2015 6:00:47 PM

 10 Bill VanPatten
You most likely said (2) sounds fine but you didn’t like (3) or (4). What
you are unconsciously experiencing in this little test is the constraints
on word formation in English. Words can change depending on what
kind of words they are. In the morphology of words, meaning the shape
of words, linguists talk about changes brought about by morphologi-
cal inflection and morphological derivation. Inflection refers to how we
add affixes—suffixes, prefixes, and what are called infixes—to words
without changing their word class (e.g., noun, verb, adjective) or their
basic meaning. Derivation refers to how we add endings or prefixes that
either change the meaning of the word or shift it from one word class to
another. (We will see examples of inflectional and derivational inflections
in a minute.) These inflectional and derivational affixes are referred to as
morphemes. Morphemes are considered the smallest sequence of sounds
that carry semantic information (some kind of meaning). For example,
the plural morpheme –s in English carries the meaning “more than one,”
as in (one) dog versus (two) dogs. The verb morpheme –d carries the
meaning “happened before the present time,” as in I watch TV (everyday)
versus I watched TV (last night). Note that by definition, the root of
any word or any simple word is a morpheme because it carries meaning.
Thus, dogs consists of two morphemes, dog and –s, as does watched,
watch and –d. Remember not to let spelling fool you. Spelling is meant to
represent underlying sounds, but unfortunately, in English, spelling and
the sound system do not form a one-to-one correspondence.
Examples of inflections include most verb endings and most noun end-
ings as well. These are suffixes. We can inflect verbs for tense (i.e., for
many languages there exists a past-tense morpheme) to produce words
like type and typed, for example. In this case, the basic meaning of type
does not change (“the act of moving one’s fingers across a keyboard to
write”); what changes is the temporality of the event (when it occurs).
Some languages inflect verbs for a good many things. Italian, for exam-
ple, inflects verbs for tense but also person and number; that is, Italian
inflects the verb depending on who or what the subject is. We would say
that Italian has person-number morphemes. Parlo thus means ‘I talk/I’m
talking,’ whereas parliamo means ‘we talk/we’re talking’ and parla means
either ‘he/she/it talks/is talking’ or ‘you-formal talk/are talking.’ Again,
the basic meaning of parl- does not change and parl- remains a verb when
it is inflected for person-number. Again, verbs can be inflected to repre-
sent a variety of meanings related to how people view events. Languages
like Chinese and English are inflectionally poor for verbs while languages
like Spanish and Russian are inflectionally rich for verbs.
Nouns inflect for plurality in many languages, as we saw in the exam-
ple on phonology. Dog refers to one dog while dogs refers to more than
one. But the basic meaning of dog does not change, nor does dog change
word class. In some languages, nouns are inflected in ways English nouns
cannot be. Quechua, spoken throughout the Andes mountains, inflects

[Link] 10 8/15/2015 6:00:47 PM

 Language 11
nouns for what English requires prepositions, and in some cases certain
kinds of determiners (e.g., articles and possessives): wasi = (the/a) house,
wasipi = in the house, wasii = my house. Spanish uses inflections on most
animate nouns to indicate biological gender: niño = small child that is
male, niña = small child that is female; profesor = male professor, pro-
fesora = female professor. Languages like Russian inflect nouns for syn-
tactic function or case (subject/nominative, object/accusative, etc.) as in
doktor znaet babušku ‘The doctor knows the grandmother’ but babuška
znaet doktora ‘The grandmother knows the doctor.’ Once again, these
inflections do not change the basic meaning of the word or its word
class. Doktor and doktora may refer to the exact same doctor in each
sentence, and both are clearly nouns. What the inflection changes is the
function or role that the noun plays in the sentence (i.e., whether it’s a
subject or not).
Certain inflections and prefixes, however, result in fundamental changes
of a word. The word either comes to mean something else or it changes
word class. In such cases, we say the word is derived from another word
(hence, morphological derivation). Let’s go back to the verb type. We can
inflect it with –ist so that we change the meaning of the word from the act
of typing to the person who does the typing. At the same time, the word
ceases being a verb and becomes a noun: I’m a terrible typist but not *Do
you typist well? This latter sentence is impossible because type has ceased
to be a verb once the –ist is added. There are other endings like –ist that
change a verb into a noun referring to the person or thing who does
something: –or as in act/actor, visit/visitor, refrigerate/refrigerator; –er, as
in beat/beater, toil/toiler, sleep/sleeper. We can take the verbs believe and
do and change them into adjectives by adding –able: believe/believable,
do/doable. In such cases, we say these words consist of two morphemes:
the root morpheme (e.g., act, visit, sleep) and the derivational morpheme
(e.g., –or/–er). Examples of other English morphemes that can be added
to verbs to make nouns include –tion (and –ation) (transport/transporta-
tion) and –ee (pay/payee). Just because words contain the same sequences
of sounds as we might find in an inflection does not mean they contain
derivational morphemes. For example, gist is not g + –ist.
Other derivational morphemes can cause changes in meaning with-
out causing a change in the class of the word. We can take the –er suf-
fix, for example, and add it to some nouns to make a new noun: labor/
laborer. We can add prefixes such as dis–, un–, and non– to adjectives
and nouns (and sometimes verbs) to create words with opposite mean-
ings: honest/dishonest, lawful/unlawful, sense/nonsense, tie/untie. Some
prefixes cause slight changes in meaning, such as re–, which we saw at
the opening of this section and which means something occurs again:
write/rewrite, build/rebuild.
In terms of morphology, a speaker of a given language possesses con-
straints on the use of morphological inflections and morphological

[Link] 11 8/15/2015 6:00:47 PM

 12 Bill VanPatten
derivations. Clearly, speakers do not allow verbal inflections to be placed
onto nouns—although many nouns can be used as verbs, at least in English—
and don’t allow nominal inflections to be placed onto verbs. When it comes
to derivations, only certain words can take un– or dis– while some words
seem to take both (or don’t sound so bad if they do). Thus, dishonest is fine
but *dislawful sounds terrible, and yet unhonest doesn’t sound so bad in
comparison, although you’d most likely never hear it.
Clearly, the morphology of words is complex, consisting of different
kinds of inflections that have different kinds of functions. The use of
these inflections is subject to constraints, such that, for example, noun
inflections can’t be attached to verbs and verbal inflections can’t be
attached to nouns. Or some inflections on nouns can only be used with
certain nouns and not others. Just like phonology and the lexicon, we
see that there is more to the nature of words than what we tend to think
of. Second language learners, like first language learners, must develop
a representation for both morphological inflections themselves and the
constraints and underlying information they contain. This is of particular
concern for those involved in the learning and teaching of what we call
“inflectionally rich” languages, such as Spanish, Italian, Russian, Turk-
ish, and others. Paradoxically, what we know about morphology and
how such inflections are presented and often practiced in language mate-
rials don’t quite mesh. That is, paradigms and lists of inflections don’t
exist in people’s minds/brains. No Spanish speaker, for example, gets to
the age of five with a paradigm for person-number endings in his or her
head that looks like this:

Singular plural
1st person -o -mos
2nd person -s -is
3rd person [nothing] -n

What is more, no second language learner winds up with such a para-

digm after much exposure to Spanish. Although this is not the place for
us to discuss language acquisition, we can say that for both first and
second language learners, person-number endings are real things, to be
sure, but they are derived after much encounter with language. That is,
a learner of Spanish comes to understand that –o refers to what we call
first-person singular only after hundreds (if not more) encounters of verbs
used in contexts in which the speaker is talking about himself or herself.
We might say that these endings are “derived” from exposure over time,
but don’t exist independently of the words with which they were learned.
For the teacher, it is enough to know that what is in textbooks and what
winds up in people’s heads aren’t the same.

[Link] 12 8/15/2015 6:00:47 PM

 Language 13

Reflection
Languages often have what are called “irregular” forms for both
nouns and verbs. For example, the plural of man is men and the
past of come is came. What do you make of so-called irregulars?
Are they derivational in the same way that we have can + s = cans
and bum + –d = bummed? Or are they something else?

Quiz
Take the following short quiz to see what you have learned so far.
Answers are given at the end, so don’t peek.

1. Phonological representation refers to . . .

a. the nature of how sounds exist in the mind of a speaker.
b. how a person makes sounds when speaking.
c. the difference between speaking casually and speaking formally.

2. Assimilation for sounds exists . . .

a. only in Spanish.
b. only in English.
c. in many languages.

3. The difference between /s/ and /z/ in ‘sue’ and ‘zoo’ suggests
that these two sounds are . . .
a. phonemes.
b. morphemes.
c. lexical items.

4. Languages vary as to whether or not they . . .

a. have vowels.
b. have open syllables.
c. allow consonant clusters at the beginnings of syllables.

5. Lemma is a fancy term for . . .

a. the meaning of a word.
b. the phonological representation of a word.
c. the underlying grammatical properties of a word.

6. Most entries in the mental lexicon contain . . .

a. semantic information (meaning).

[Link] 13 8/15/2015 6:00:47 PM

 14 Bill VanPatten

b. grammatical properties.
c. both a and b.

7. The arguments of verbs are related to semantics and meaning

but have direct impact on . . .
a. grammatical properties of a sentence.
b. phonetic realization of the verb.
c. connections with other words.

8. Why is do a bit odd compared to other verbs such as seem,

steal, and catch?
a. It only appears in the present tense.
b. It only has one argument.
c. It lacks semantic information (meaning).

9. Which of the following illustrates an inflectional morpheme as

opposed to a derivational morpheme?
a. do + –s = does
b. un + even = uneven
c. transport + ation = transportation

10. A critical aspect of morphemes is that they . . .

a. lack semantic information.
b. carry meaning.
c. are free to attach to any word.

[Answers: 1. a; 2. c; 3. a; 4. c; 5. b; 6. c; 7. a; 8. c; 9. a; 10. b]

Syntax
Words combine together to form phrases and sentences. There are noun
phrases (the dog, a dog, my three dogs, the man with three dogs), verb
phrases (eats, eats the pie, eats a lot, eats pies almost every day), preposi-
tional phrases (at home, in the backyard), and so on. These phrases work
together to make sentences: The man with three dogs eats pies almost
every day at home. How phrases form and how they work together to
make sentences is determined by syntax. A speaker’s syntactic competence
sets constraints on the nature of phrases and sentences. It is the compo-
nent of language that governs structural relationships among words. This
competence is abstract and not easy to describe. For example, most of
us know what the subject of a sentence is, right? The average person
would recite a high school grammar rule: the subject of the sentence is

[Link] 14 8/15/2015 6:00:47 PM

 Language 15
the person or thing that performs the action. This is a fine description
for action verbs such as cuts, as in Robert always cuts his meat into little
pieces. Robert is a cutter. But what about the verb seems? What is the
action in seems? There is none, and we described seem earlier as an event
where X experiences Y. X is an experiencer and not an actor/agent of an
action and yet we know that Robert is the subject of Robert seems sad.
But we can’t say Robert is a seemer. What about the verb is? Again, no
action there, yet we know that the dog is the subject of The dog is asleep.
Once more, the dog is not an iser or a be-er. Passive sentences also pose
a problem for any kind of semantic definition of subject. In The cow
was kicked by the horse we know that the horse performed the action
of kicking, and yet we also know that cow is the subject of the sentence.
And then there is the issue of “person or thing.” What is a thing? In the
sentence That John believes he can talk to animals makes me wonder if
he’s gone mad, what is the subject of makes? In this case, it’s an entire
clause (that John believes he can talk to animals). Is this a thing like wine,
knife, or sun? What’s the thing represented by it (subject) in It’s raining
or It’s almost three o’clock?
What these examples demonstrate is that a speaker has an abstract
representation for “subject of a sentence” that defies any easy definition
that refers to typical real-world events and meanings. The nature of sub-
ject involves a special (and once again, abstract) structural relationship
to a verb. This structural relationship is not necessarily linear, either. For
example, the subject is not “the thing that comes before the verb.” In
Mary doesn’t eat shellfish, doesn’t is not the subject of eat. And what
happens when we have compound verbs: Has anyone seen my laptop? In
this case the verbal unit has seen has been split by a noun. Is that noun
(anyone) the subject because it comes before the verb seen or is it not the
subject because it comes after the inflected verb has? In languages like
Japanese, two nouns appear before the verb, subject and object, so that
standard word order in Japanese is subject-object-verb or SOV. In this
case, more than one thing appears before the verb. If you try to say that
the first noun in the sentence is the subject, then aside from the fact that
not all subjects are typical nouns (To err is human, to forgive divine),
some structures in some languages require that the first noun normally
not be the subject, as in Spanish gustar constructions: A María le gusta
mucho Juan (lit: to Mary is very pleasing John). Spanish has no verb that
means to like and indirect object-verb-subject word order is standard with
gustar. And although subject-verb-object and subject-object-verb are the
most frequent word orders in the world, object-verb-subject languages do
exist (e.g., Hauarijio, a Uto-Aztecan language spoken in parts of north-
western Mexico).
In some languages, word order does not obligate a subject to appear any-
where in front of the verb, and in some cases, the syntax may prefer or
require the subject to appear postverbally. Spanish is a good example of such

[Link] 15 8/15/2015 6:00:47 PM

 16 Bill VanPatten
a language. Standard question formation is one example in which Spanish
typically follows verb-subject (VS) or object-verb-subject (OVS) order:

(5) ¿Ya habló Roberto?

already spoke Robert?
‘Did Robert already speak?’

(6) ¿A quién vio María?

ACC–who saw Mary?2
‘Who did Mary see?’

(7) Después que regresó Juan, Pablo pudo irse.

after returned John, Paul was able to leave.
‘After John returned, Paul was able to leave.’

What is more, languages like Spanish and Chinese may omit subjects
and subject pronouns altogether, as in the examples below.

(8) Vino/no vino.

came/no came
‘He/She/It came/didn’t come.’

(9) huí jiā le

already gone home
‘He/She has already gone home.’

(10) chī b o le
already full
‘I/He/She am/is already full.’
These are called null subject languages because the actual expression of
a subject does not appear in the sentence. It is null. If there is no subject
expressed, where is the linear relationship to the verb?
So linearity is not a way to describe subjects in a language, even though
in English the description might frequently work. What we want is a defi-
nition of subject that fits all languages, not just English. Unfortunately,
that definition is too complex for this introduction to syntax and requires
specialized knowledge about the abstract nature of sentences. It’s enough
to know here that you have this abstract knowledge and didn’t even
know it. It is implicit knowledge as described in the outset of this module.
Let’s examine something different. English allows adverbs in a variety
of positions, but prohibits them between verb and complement:

(11) John watches TV often.

(12) John often watches TV.
(13) *John watches often TV.

[Link] 16 8/15/2015 6:00:47 PM

 Language 17
French, however, allows these same adverbs to appear between verb
and complement but disallows them between subject and verb.

(14) Jean vois la télé souvent.

(15) Jean vois souvent la télé.
(16) *Jean souvent vois la télé.

What is the difference between French and English here? And why
does Spanish look like French but not English? How can we explain
these differences? Every native speaker of English, Spanish, and French
“knows” which of these sentences are good and which ones aren’t. This
is part of their syntactic competence. As we will see later, syntactic com-
petence should not be confused with knowing the rules of grammar or
good speech. We will touch on that issue a bit later.
As one more example, let’s go back to auxiliary do for a moment. One use
of this verb is in question formation: Does Will know Grace? Why is this
do necessary? It’s not necessary in languages like Spanish, Turkish, Chinese,
or Japanese. In Spanish, for example, such a construction doesn’t exist, but
what does exist is subject-verb inversion to ask basic yes/no questions: ¿Con-
oce Will a Grace? In English, the equivalent is just not possible: *Knows Will
Grace? (although this kind of structure was permissible in Old English along
with rich verbal inflection for person and number). What abstract properties
of English require a do to make simple questions and are these related to the
same properties that allow Spanish to use subject-verb inversion? In current
linguistic theory, we do find they are related to singular abstract underlying
features of syntax that work one way in English and another in Spanish.
Interestingly, these features are related to those that allow English to per-
mit wh– words like what, when, and where in two positions (Where is Bill
going?/Bill is going where?) but in Chinese, only in one position (Bill go(es)
where?/*Where Bill go(es)?). In short, underlying syntactic competence is
much more abstract and much more complex than what textbook rules and
elementary school rules of thumb would lead us to believe.

Reflection
Look at the following sentences. What do they tell you about the
syntactic competence of speakers of English regarding two-place
verbs (e.g., look up, turn on/off, take in)? Remember that an aster-
isk means the sentence is disallowed.

(1) Dave looks up the word in the dictionary.

(2) Dave looks the word up in the dictionary.
(3) Dave looks it up in the dictionary.
(4) *Dave looks up it in the dictionary.

[Link] 17 8/15/2015 6:00:47 PM

 18 Bill VanPatten
But syntax isn’t just about what is allowed and disallowed, or how
we define structural relationships. Syntactic competence also involves the
knowledge that sentences can be ambiguous. Look at the following two
sentences:

(17) Mikey saw Ronny after he came home from France.

(18) Juliet told Susan she broke her arm last night.

What would you respond if you were asked the following question
about (17): Who came home from France? In this particular instance,
the answer could be either Mikey or Robert, even if you have a pref-
erence. The pronoun he can take either antecedent, Mikey or Ronny.
Your syntactic competence allows this, which results in ambiguity. For
sentence (18), two interpretations are also possible: Juliet spoke to Susan
last night and said she broke her arm, or Juliet broke her arm last night
and then told Susan. Because humans possess syntactic competence, they
know when a sentence can be ambiguous. People may build up prefer-
ences for one kind of interpretation over another during language acqui-
sition and throughout life, but sentences such as these are ambiguous
in and of themselves. Ambiguity has to do with syntactic relations, or
how one element in a sentence relates to another in terms of underlying
structure.
Like native speakers, second language learners must develop the same
kinds of “intuitions” about sentence structure. A great deal of research
suggests that this is indeed what happens. However, as you will learn
elsewhere, some parts of syntactic competence are acquired quickly, oth-
ers slowly, and others even more slowly. The takeaway point for lan-
guage teachers here is that syntactic competence in L2 learners does not
look like textbook rules for how sentences are formed. We will touch on
this later in this module.

Pragmatics
Imagine you’re sitting in a meeting. Things are getting heated as the
assembled members argue their points passionately. One person makes
a snide remark to another. The chair of the meeting stands up and says,
“Why don’t we take a break?” Which of the following makes the most
sense?

A. You say out loud: “Because we don’t have time. It’s already 2:00 and
there’s more to do. What’s more, Harry isn’t done making his point.
There are also ten more items on the agenda.”
B. You and several others mutter: “Good idea.”

While both responses are possible, more than likely B is what would
happen. Why? Because the response in A is an answer to an open ended

[Link] 18 8/15/2015 6:00:47 PM

 Language 19
question, as in “Why, with the Earth spinning as fast as it does, don’t we
all fly off the planet?” “Because of gravity.” Whoever answers A takes
the chair’s utterance to be a literal question. The other response, B, indi-
cates that people take the chair’s utterance as a recommendation, if not a
command, something equivalent to “OK. We’re taking a break. We need
some air before continuing.”
Response B demonstrates those people’s pragmatic competence, refer-
ring to how people interpret language in context. The field of study called
pragmatics within linguistics aims to explain how the factors outside
of language itself (e.g., phonology, syntax, morphology, lexicon) affect
meaning, both literal and non-literal. In linguistics, we might say an
utterance is a declarative, a question, or a command. But in reality, not
all utterances with the shape of a declarative are a declarative. Not all
utterances with the shape of a question are a question. And not all utter-
ances shaped like commands are commands. What is more, a declarative
may be interpreted as a declarative but with an intent overlaying the
actual words, as indicated by the other person’s response. Examine the
following interchange:

JOE: I need ten bucks.

HARRY: Forget it.
JOE: But it’s really important. I’m in a jam.

Joe’s declarative utterance is not just a description of his needs, it’s

actually a request. Overlaying the words is “Can I borrow ten dollars?”
Harry’s “command” does not literally mean that Joe should suddenly not
remember something; his response is a refusal. He means he won’t lend
him the money, and yet doesn’t need to say “I won’t lend you the money.”
That is understood by Joe, as indicated in his response.
What these examples demonstrate is that conversations are not just sets
of utterances strung together. Conversations are also sets of speech acts.
Speech acts refer to such things as requesting, recommending, apologiz-
ing, insulting, complimenting, and so on. Some are direct and some are
indirect. Direct speech acts are those in which the words of the utterance
closely match the speaker’s intent, as in “Answer the phone, please.” This
utterance is a request and is very direct. Indirect speech acts are like those
we saw earlier. “Why don’t we take a break?” is not a question. It’s an
indirect speech act that is a recommendation. And instead of “Answer
the phone, please,” someone could perform an indirect speech act to
make the same request by saying, “Do you mind?” This is not a literal
question. One final example can be found easily in compliments. “That’s
a nice hat” is a direct speech act. “Where’d you get that hat?” could be
an indirect speech act, and it would not be unreasonable for someone to
say “Why, thank you. I can get you one if you want.”
Irony and jokes often involve culturally specific pragmatics. At a con-
ference on applied linguistics a few years back, a speaker was talking

[Link] 19 8/15/2015 6:00:47 PM

 20 Bill VanPatten
about gay humor and related this joke to the audience: What does a
lesbian bring on a second date? A U-Haul. What does a gay man bring
on a second date? Another date. The answers to each question are not
meant to be taken literally. In the first example, the idea is that lesbians
are quick to couple and move in together, not that a lesbian actually
brings a U-Haul trailer packed with her belongings on the second date.
In the second joke, the idea is that gay men are not quick to couple and
like to “shop around.” A gay man really doesn’t bring another person
with him on his second date with someone else. Note that the jokes work
best when told back to back and involve culturally specific information.
At the applied linguistics conference where these jokes were reviewed, a
number of native speakers in the audience did not get the jokes because,
not being part of a sub-culture within the U.S., they did not possess the
cultural knowledge to understand the humor.
As people’s competence in the purely formal areas of language devel-
ops (e.g., phonology, syntax, lexicon), their competence with speech acts
also develops. The result is an underlying pragmatic competence that is
evident in how language is used. Just like syntax, phonology, and other
aspects of language, pragmatic competence can vary from language to
language (or better yet, culture to culture). A native Brazilian who was
studying for a doctoral degree in the U.S. once told me that when she first
arrived, she had a number of conversation breakdowns when someone
used a Why don’t you . . . utterance as a recommendation. In her Portu-
guese, the use of why to make a recommendation in this way didn’t exist,
and she kept misinterpreting people who’d use this particular indirect
speech act. She’d often give long explanations to answer the why, often
being met with strange looks. She eventually made pragmatics her focus
of research.

Reflection
Examine the interchanges and then respond as though you were
the person in the situation. Do you respond the same each time?
Explain your responses in terms of the speech acts you’re respond-
ing to.

PHILLIP: I don’t have much experience with women.

JEAN: As good looking as you are?
PHILLIP: Yes.
JEAN: Get out of here!
PHILLIP: ________________

MARY: They’re after us!

SUSAN: But I don’t see them.

[Link] 20 8/15/2015 6:00:47 PM

 Language 21
MARY: I do. Let me take care of it.
SUSAN: But I can’t leave you behind!
MARY: Go! Get out of here!
SUSAN: _________________________

More on Language as an Implicit Knowledge System

(and One More Look at Complexity and Abstractness)
Psychologists often talk about explicit and implicit knowledge systems.
Explicit knowledge systems are those that a person can verbalize in some
way and is aware of. Conscious knowledge is another way of talking
about explicit knowledge. Say, for example, a person recites a telephone
number to someone. That telephone number exists as explicit knowledge
that can be called forth. In language, a person might know something
like the conjugations of a verb or have a working definition of what a
reflexive verb is. This knowledge is explicit because that person can tell
you in one way or another what those things are, how they work. We
might say a person has explicit or conscious rules about language. This
does not mean the person is correct, only that he or she has some explicit
knowledge he or she thinks is correct. If we review the discussion on the
definition of “subject” in the section on syntax, we see that some people
have conscious knowledge about what they think a subject is, but that
knowledge is far from adequate or even accurate.
The kind of mental representation for language we are discussing in this
module, however, is not explicit. It is implicit—and very often abstract,
as we saw with some examples from syntax and morphology. By implicit
we mean that the knowledge exists outside of one’s conscious awareness,
that it is largely unavailable for description or articulation—except after
much introspection and, even then, this is the job of linguists and the
average person can never do this without training (e.g., Have you, the
reader, arrived at a definition for “subject of sentence” yet? Have you fig-
ured out why re– can combine with some verbs and not others?). A per-
son can be aware of having implicit knowledge. For example, the reader
knows that he or she has English in his or her head, that it exists in some
form or another in the mind/brain. Again, what makes this knowledge
implicit is not that we don’t know we have it but that we can’t articulate
its nature.
The problem in discussing and researching second language acquisition
is that learners may possess both kinds of knowledge, explicit and implicit,
but only the latter is of interest to us. Let’s take the case of asking ques-
tions. The language learner may be able to say, “to make a yes/no question
in English, put an appropriate form of do in front of the sentence”: You
know John  Do you know John? He knows John  Does he know
John? But for the advanced speaker/knower of English, this is not how yes/

[Link] 21 8/15/2015 6:00:47 PM

 22 Bill VanPatten
no questions are represented in mental representation. Instead, there are
particular syntactic operations that must be satisfied and are related to the
same reason why Spanish can invert subject and verb to make questions
but English cannot: Juan conoce a Pedro  
¿Conoce Juan a Pedro? John
knows Peter  *Knows John Peter? To drive the point home, the explicit
rule is one thing, the implicit stuff of mental representation is quite another.
The two are distinct entities and generally do not resemble each other.
If you are a typically curious reader, then you may be asking, “Just
how far off is the rule of thumb about making yes/no questions from
what actually exists in mental representation?” The answer is “quite far
off.” To illustrate, let’s examine current linguistic theory about the nature
of yes/no questions in languages like English and Spanish. To begin, we
need to remind ourselves about two things regarding language. The first
is that mental representation includes a lexicon, the component of long-
term memory where information about words is stored. The reader will
remember that we said all entries in this lexicon encode not just meaning
(where there is meaning to encode) but also grammatical information.
For English and Spanish, some example items appear here with repre-
sentative information. That is, we are not listing all of the information
encoded within the lexical entry. (Recall from note 1 that features are
binary, so they carry + and – symbols. The + symbol means the item is
“positive” for that entry, while the – symbol means it is not.)

do: <+V>, <-N>, <+T>, <+present>, <-past>

does: <+V>, <-N>, <+T>, <+present>, <-past>, <+3rd person>, <+sing>
did: <+V>, <-N>, <+T>, <-present>, <+past>
will: <+V>, <-N>, <+T>, <-present>, <-past>
drink: <‘drink’>, <+V>, <-N>
drinks: <‘drink’>, <+V>, <-N>, <+present>, <-past>, <3rd person>, <sing>
tomo: <‘drink’>, <+V>, <-N>, <+T>, <+present>, <-past>, <1st per-
son>, <sing>
tomas: <‘drink’>, <+V>, <-N>, <+T>, <+present>, <-past>, <2nd
person>, <sing>
toman: <‘drink’>, <+V>, <-N>, <+T>, <+present>, <-past>, <3rd
person>, <pl>

Thus, the verb do, for example, has no meaning but has the gram-
matical features listed under each of its entries. Drink has meaning (not
spelled out above but shorted-handed as ‘drink’) and the grammatical
features indicated. Spanish has no equivalent to do but does have entries
for the equivalent of drink. What the reader may notice is that the various
entries for drink and for toma differ in that the latter contain a feature
labeled ‘T’ whereas the former do not. T stands for tense. This means
that Spanish verbs carry abstract tense features whereas English verbs do
not (except for do and other auxiliaries such as have). This T feature will
come in handy in a moment.

[Link] 22 8/15/2015 6:00:47 PM

 Language 23
Now, the second thing we need to remind ourselves is that sen-
tences consist of internal hierarchical phrases: verb phrase (VP), noun
phase (VP), prepositional phrase (PP), and so on. These are called lexi-
cal phrases. What the reader may not know is that sentences also have
functional phrases, of which the two major ones are the complementizer
phrase (CP) and the tense phrase (TP). The CP is basically the “root”
phrase of any sentence or clause. The TP is where features related to
tense, verb agreement, and other things happen. All lexical items are
generated in lexical phrases only; the functional phrases participate in
grammatical operations, as we will see shortly. What is more, the CP and
TP are “higher up” in a sentence compared to lexical phrases. The base
structure looks like this (ignore Spec for now):

Words are inserted when a yes/no question is required in English or

Spanish. Like this:

Again, note that the lexical items from our mental dictionary are inserted
only into lexical phrases. But the above clearly aren’t yes/no questions
when we say each sentence out loud in a linear fashion. Instead, they look
like declaratives: Bill drinks vodka and Bill toma vodka. Here’s where the

[Link] 23 8/15/2015 6:00:47 PM

 24 Bill VanPatten
T feature enters into the picture. Questions have a hidden T feature in the
CP under the C node. They also have a T feature under the TP in the T
node. These features are looking for something for them to “be complete.”
Otherwise, they are just hanging around and feeling unfulfilled. Spanish
verbs carry this T feature. So, verbs in Spanish are “attracted” up to the T
phrase so that the TP can be satisfied. Then the verb is attracted once more,
but this time to the CP so that the T feature there can be satisfied. This is
how it looks (ignore what happens to the subject Bill for now):

And this is how we get verb-subject word order for yes/no questions in
Spanish. Now, for English, let’s remind ourselves of a crucial feature of
main verbs: they don’t carry a T feature. Thus, they can’t participate in
the same kind of “movement” and “attraction” that Spanish verbs can.
So, what does English do? It inserts a dummy verb, do, with T features
into the TP to satisfy the T feature of this phrase. Then, the dummy verb
is attracted up to the CP the same way Spanish main verbs are. Here is
what the English question looks like:

[Link] 24 8/15/2015 6:00:48 PM

 Language 25
It is this complex interplay of units with syntactic computation that
yields Does Bill drink vodka?, while simultaneously ruling out Drinks
Bill vodka? It is this complex interplay of units that yields ¿Toma Bill
vodka? in Spanish while ruling out the need for a dummy verb. In short,
both Spanish and English must solve the “T problem” of yes/no ques-
tions; they simply have different devices for doing so.
The reader with excellent memory may recall at this point that earlier
we said Middle English made yes/no questions like Spanish. You might
guess that the reason for this is that verbs in Old English had underlying
T features (and were morphologically rich compared to verbs in Modern
English). So, the declarative in Old English would have looked something
like Bill drinketh vodka while the yes/no question would have looked
like Drinketh Bill vodka? And the abstract representation showing move-
ment of the verb to satisfy the “T requirements” of TP and CP in Old
English would have looked strikingly similar to what we showed above
for Spanish:

The point of this brief exposé is to illustrate the underlying complexity

and abstractness of something as common as yes/no questions, as well
as to highlight that all of this knowledge is in the minds/brains of speak-
ers of Spanish and English. They just don’t know it’s there. It is implicit
knowledge. And it is this implicit knowledge that is the object of second
language acquisition.

On Textbook Grammars and Prescriptive Grammars

An often misunderstood notion by second language teachers and learners
alike is that the grammar they learn in textbooks is either what exists in a
native speaker’s head or is what the language learner has to “internalize.”
Nothing could be further from the truth. Textbooks provide pedagogi-
cal grammars, rules of thumb to help students bootstrap themselves into

[Link] 25 8/15/2015 6:00:48 PM

 26 Bill VanPatten
learning, but these rules do not “turn into” the mental representation
that underlies language. Let’s examine a few concrete examples.
We touched upon yes/no questions made with do in English previously.
A textbook rule would be something like “insert do and invert it with
the subject of the sentence,” which on the surface seems to capture what
happens. But what actually exists in the mental grammar of a speaker
of English? As we also saw previously, there is something much more
abstract and complex than this simplistic pedagogical rule. And if you
reflect on it for a moment, how is it possible for something like “insert
do and invert it with the subject” to ever become this abstract, complex
interaction of features and syntactic operations that we saw earlier? The
answer is, it can’t. Pedagogical rules are one thing, and underlying mental
representation is another. Pedagogical rules are a short-hand way to talk
about something that is too complex to describe to the average person/
student. What is more, most, if not all, pedagogical rules are like this.
That is, they don’t necessarily correspond to underlying representation.
To be sure, this discussion is not a suggestion that pedagogical gram-
mars ought to be abandoned, that they are not useful, that they serve
no purpose. This is the stuff of empirical research in second language
acquisition. But to confuse the rules that learners are given in classrooms
and textbooks with what winds up in the mental representation would
be wrong. To think that such rules “turn into” the mental representation
(that is, are somehow transformed) would be equally wrong.
Somewhat related to the notion of pedagogical grammars is that of
prescriptive grammars. Prescriptive grammars are those that exist for
native speakers, to instruct them in the “proper use of grammar” because
people may be judged by their “bad grammar.” A classic example is ain’t.
This negative is frowned upon in certain circles of English users. The child
who says I ain’t got none might be corrected by his parent or teacher with
“Don’t say that. Say ‘I don’t have any’.” Similarly, people might be cor-
rected for who and whom, me and I, and other things in English. They
are aspects of prescriptive grammar related to speech communities and
what people value. They are not related to mental representation and
underlying competence. Here’s what we mean. In an English speaker’s
mental representation of language, the use of ain’t is constrained such
that there are permissible and disallowed sentences. I ain’t got none is a
permissible use of ain’t. But I ain’t have any sounds weird, at best. What
is more, no native speaker would break up ain’t into ai and not the way
he or she could with don’t/do not and haven’t/have not, for example.
Let’s examine several interchanges:

Scenario A
STEVE: I don’t have any.
BOB: Oh please. Gimme some.
STEVE: Listen. (slowing his speech) I do not have any. None. Nada.

[Link] 26 8/15/2015 6:00:48 PM

 Language 27
Scenario B
STEVE: I ain’t got none.
BOB: Oh, please. Gimme some.
STEVE: You got wax in your ears? (slowing down) I ai not got none.

Scenario A contains language that is completely permissible by the

mental representation of English. Scenario B does not. (The last sentence
should have an asterisk to show it is not allowed in English.) And any
speaker of English knows this, regardless of whether ain’t is “proper
English” or isn’t.
In other cases, the term grammar is used in everyday circles to refer to
lexical choice and sometimes even punctuation. A book that hit the best-
seller list in the United States and the United Kingdom was called Eats,
Shoots, and Leaves. The author contrasts the title with “Eats shoots and
leaves.” The first is a series of verbs while the latter is a description of what
an animal eats. The author claimed that the inappropriate use of punc-
tuation or the lack of punctuation is part of “bad grammar.”
When we use the term grammar in linguistics, we are not talking about
prescriptive grammar. We are not talking about punctuation, dangling
participles, split infinitives, and all those other things that teachers like
to teach their students about in elementary and secondary schools. When
we say “the grammar in the English speaker’s mind,” we mean the under-
lying mental representation or competence. We do not intend to mean
how people think we should talk, but instead what in our competence
allows or disallows various structures. Linguists even use the term gram-
mar to talk about the specific properties in that mental representation,
such as the grammar of verb movement (as in Spanish yes/no questions),
or the grammar that governs ain’t. At the same time, when linguists talk
about grammars, they do not mean pedagogical grammars. They do not
mean the lists of verb forms, lists of rules, and all the other stuff we find
in language textbooks or those self-help books on “how to brush up your
grammar and impress others.” In short, grammar refers to the idea of a
mental representation, a competence consisting of implicit and abstract
knowledge and constraints, including constraints on the acceptable use
of so-called bad grammar.

Reflection
Someone says Me and John are going to the store. You need any-
thing? That person is a college educated business-person. Some-
one next to you whispers under her breath, “John and I.” What
does the first person’s utterance tell you about the English speaker’s
mental representation of compound subjects with a pronoun and
what does the second person’s reaction say about prescriptive gram-
mars? (Note: The first person in our scenario—or any other native

[Link] 27 8/15/2015 6:00:48 PM

 28 Bill VanPatten
English speaker—would never say *Me am going to the store. The
vacillation between I/me, him/he, and so on, only occurs in com-
pound subjects such as Me and him went and in compound objects
such as He told John and I/me to leave.)

Quiz
Take the following short quiz to see what you have learned since
the last quiz. Answers are given at the end, so don’t peek.

1. Syntactic competence involves . . .

a. constraints on the nature of phrases.
b. constraints on the nature of sentences.
c. both a and b.

2. Syntactic competence also involves . . .

a. the relationship of words to each other.
b. word order, such as verb-object and object-verb.
c. both a and b.

3. Two adjectives that describe syntactic competence are . . .

a. abstract and implicit.
b. prescriptive and pedagogical.
c. meaningful and ordered.

4. Which of the following concepts best helps us understand prag-

matics and pragmatic competence?
a. meanings of words
b. speech acts
c. declarative sentence

5. Which of the following influences pragmatics but not syntax?

a. argument structure of verbs
b. cultural background
c. word order

6. Implicit knowledge of language means that people . . .

a. are not aware of their knowledge.
b. can’t really articulate what the knowledge is.
c. both a and b.

[Link] 28 8/15/2015 6:00:48 PM

 Language 29

7. Which of the following is true about pedagogical grammars

and pedagogical rules?
a. They are accurate reflections of mental representation.
b. They may help learners bootstrap themselves into the lin-
guistic system.
c. They can, with practice, become underlying competence.

8. Which of the following is most likely true about prescriptive

grammars based on what you have read?
a. They describe how people actually use language.
b. They are used by linguists to understand mental representa-
tion.
c. They are socially and culturally motivated.

[Answers: 1. c; 2. c; 3. a; 4. b; 5. b; 6. b; 7. b; 8. c]

Other Perspectives on Language

The perspective taken on language and representation in this module
is one of several. While all linguists deal with notions such as syntax,
morphology, nouns, verbs, constraints, and so on, they may talk about
them in different ways. In fact, there are probably about four or five
theories or frameworks for mental representation that exist out there.
There are others, however, that do not take the perspective that there is
an underlying mental representation for language. These kinds of per-
spectives are taken largely by non-linguists (e.g., psychologists), some
of whom see language not in its abstract sense but as a network of
“stuff” in the mind/brain that yields language-like behavior. This mod-
ule is not a place to enter into the debate on the exact nature of what
underlies what we call “language,” and, what is more, there is not a fact
about language presented in the present module that any framework
or theory would deny. That is, no matter what one’s perspective is,
there is a sound system and sounds behave in certain ways. There is a
lexicon and words have properties. There is a syntactic component such
that there are allowed and disallowed sentences, and people know such
things if you ask them (the right way). There is a pragmatic domain by
which people link sentence structures with speech acts. And all of this
stuff is implicitly stored in the mind/brain. What is debated is how such
knowledge originates and is learned by people, and the degree to which
first and second language acquisition are similar or different. That is
the topic for a different discussion.

[Link] 29 8/15/2015 6:00:48 PM

 30 Bill VanPatten
Discussion Questions and Projects
1. Phonemes are said to be “contrastive”; that is, if you substitute one
sound for another, you get a different word. This is clear with the
sounds /b/ and /v/, for example: buy is not vie, boat is not vote, and so
on. What about /s/ and /z/? Give examples where the use of /s/ vs. /z/
results in a meaning change and then compare this to the example of
plural formation in English discussed early in this chapter. Is the differ-
ence between /s/ and /z/ different from the difference between /b/ and
/v/, for example?
2. As we saw, a morpheme is the smallest sequence of sounds that carries
meaning, as in talks where –s means “someone other than the speaker
and the listener” and presidential where –al means “having the char-
acteristics of.” What morphemes can you isolate in these words and
what do they mean? Include the roots of words: (1) decode/debone/
devaluation; (2) who’s/whom/whose; (3) refrigerator/refrigerating/frigid.
3. Morphemes are said to be the smallest sequences of sound that carry
meaning. Does a morpheme always carry the same meaning? Look
over these words and decide what –ette means.
kitchen/kitchenette
flannel/flannelette
usher/usherette
4. See if you can determine why the asterisked sentences aren’t possible
in English. [Hint: think about your mental lexicon, not your syntac-
tic competence.]
(1) John disappeared.
(2) I made John disappear.
(3) *I disappeared John.
(4) Mary died.
(5) I made Mary die.
(6) *I died Mary.
5. Prepare a test for some friends who are speakers of English using the
sentences below. What do you find after sampling ten or more people in
terms of their syntactic competence? Are they pretty much in agreement?
(1) Bobby likes to look at himself.
himself can refer to Bobby yes no
himself can refer to someone else yes no
(2) Bobby told John to look at himself.
himself can refer to Bobby yes no
himself can refer to John yes no
himself can refer to either Bobby or John yes no
himself can refer to someone else yes no

[Link] 30 8/15/2015 6:00:48 PM

 Language 31
(3) Bobby told John that he likes to look at himself.
himself can refer to Bobby yes no
himself can refer to John yes no
himself can refer to either Bobby or John yes no
himself can refer to someone else yes no
(4) Bobby remarked on how John likes to look at himself.
himself can refer to Bobby yes no
himself can refer to John yes no
himself can refer to either Bobby or John yes no
himself can refer to someone else yes no
6. Most practicing teachers believe there is a relationship between what
is taught and practiced (e.g., textbook rules) and what learners learn.
But this module suggests otherwise. Why do you think so many
teachers believe in this relationship? Have your own thoughts on this
matter changed after working with this module?

Notes
1. In linguistics, features (properties) are binary: they are there or they aren’t.
Linguists use the + sign to indicate the feature is there and the – sign when
it isn’t. So, <+N> <−V> means that the word has “noun properties and not
verbal properties.”
2. ACC stands for “accusative,” meaning the direct object of the verb.

Suggested Further Reading

Language as Representation
Akmajian, A., Demers, R. A., Farmer, A. K., & Harnish, R. M. (1995). Linguis-
tics: An introduction to language and communication. 4th ed. Cambridge,
MA: MIT Press.
Fasold, R. W., & Connor-Linton, J. (2006). An introduction to language and lin-
guistics. Cambridge, UK: Cambridge University Press.
Pinker, S. (2007). The language instinct. New York: Harper Collins. [See espe-
cially the chapters “Mentalese,” “How Language Works,” and “Words, Words,
Words.”]

[Link] 31 8/15/2015 6:00:48 PM