Mulroy

The War Against Grammar

by David Mulroy
I practice a dying profession. I have been teaching Classics (ancient
Greek and Latin) at UWM since 1973. That year, I joined a department
with five other tenure-track faculty members. Since then, no new
appointments have been made. Classics has been merged into a
Department of Foreign Languages and Linguistics with other struggling
programs (German, Hebrew Studies, Slavic Languages and Linguistics).
Four of the original six classicists remain, but one will retire this year. Of the
other three, two of us are in our mid-fifties; the third, a youthful sixtysomething. When we go, there is little chance that any of us will be
replaced. And the situation at UWM is typical of Classics programs
nationwide. Enrollments started to slide in the mid-sixties and continue to
do so. The number of Classics majors dropped by 30 per-cent between
1971 and 91; in 1995, over a million B.A. degrees were awarded; six
hundred in Classics. 1 The handwriting is on the wall. How come?
At the heart of Classics are courses in the Greek and Latin languages.
These days hardly any students take Latin; even fewer take Greek, which
is harder. This is not for lack of initial interest. Every September at UWM
twenty to thirty eager students sign up for beginning Latin; ten or so for
Greek. If a third of them continued to study these subjects for four or more
semesters as most of them plan to do we would have a flourishing
program. Instead we lose virtually all of our prospective language
students by the end of the first year.
You may think that we are just bad teachers. Our experience, however, is
typical of Classics programs everywhere. The fatal problem is this. In order

Mulroy

to learn Latin or Greek, students need to understand English grammar.

These days very few American college students do.
The problem is not limited to Classics. Many foreign language programs
are struggling. And enrollment statistics do not tell the whole story. In order
to survive, foreign language teachers increasingly rely on
communicative rather than grammatical syllabi. They try to immerse
their students in the life styles of the people who speak the target
language in the hope that their students will pick up the language
effortlessly, the way they learned English. You cant miss this trend in
colleges and high schools. A flyer advertising beginning Japanese at
UWM emphasizes the fact that students will learn origami, the art of paper
folding. Shortly before I switched my son to home schooling, a major
assignment in his French class consisted of frying sliced mangoes,
allegedly a popular treat in Francophone Africa. It even happens in Latin
classes. An upbeat New York Times article on the alleged recovery of
Latin in some secondary schools features a class that uses dry ice to
recreate Virgils underworld. You get the picture.
I learned about grammar in a parochial school from no-nonsense (and
no-dry ice) nuns who had me diagraming complicated sentences in the
fourth grade. This gave me an understanding of the structure of language
that is by far my most valuable intellectual possession: every word, one of
eight parts of speech; every clause analyzable into one of four basic
structures.
It took me a long time to realize that my students did not share this
knowledge, which I had assumed to be the common possession of all
elementary or grammar school graduates. I was not surprised when
students needed to review the difference between participles and

Mulroy

gerunds, but the discovery that they often could not distinguish between
a noun and a verb or identify the grammatical subject of a sentence left
me incredulous. In the late seventies, an outspoken student who became
a close friend pulled the wool from eyes. In response to some patronizing
remark of mine, he said: You teachers are always putting us students
down for not understanding grammar, but you have never taught us
grammar. Maybe if you taught it, we would understand it.
From then on I always asked my students about their training in grammar
and the truth of these comments was constantly confirmed. My students
never had to master the fundamentals of grammar. The basic concepts
were presented briefly, if at all, and with evident distaste on the part of
their teachers. As a result, I added two weeks of English grammar review
to my elementary language classes, and met with some success. As time
has passed, however, student ignorance of grammar has deepened. I
once expected to salvage a half a dozen Latin students from a group of
twenty-five beginners. Now I am happy with two.
In 1996 the Wisconsin Department of Public Instruction was drafting a new
set of standards for K-12 education. A draft was published with a request
for comments. The standards included no reference to training in
grammar. I attended public hearings and wrote an opinion piece for the
Milwaukee Journal Sentinel urging the inclusion in the standards of the
ability to identify parts of speech and diagram sentences. I expected
such an obviously sound suggestion to be embraced enthusiastically by
the states educational establishment. How could any sane person object
to the proposition that high school graduates should know the parts of
speech and understand the structure of sentences? When I was asked to
serve on a subcommittee working on the language arts standards, I
agreed to do so with high hopes.

Mulroy

According to E. D. Hirsch of Cultural Literacy fame, the nations

educational establishment occupies its own Thoughtworld, in which
they are secure from all criticism. This was my first encounter with the
Thoughtworld. I found that one of its features is the arbitrary redefinition of
words. The case in point is the word standard. Unless you live in the
Thoughtworld, you think that an intellectual standard involves a specific
level of ability or knowledge that a person must display to obtain a given
distinction like a high school diploma. In another context, employers might
require prospective secretaries to type eighty words a minute. That would
be a standard, right?
In the Thoughtworld, the situation is far more complicated. Here, it turns
out, there are three kinds of standard. There is the content standard,
which specifies the area in which ones ability is assessed; the
performance standard, which describes how ability is assessed, and
finally the proficiency standard, which sets the specific level of
competence that one must display. A content standard for hiring a
secretary might state that a candidate should be tested in word
reproduction, while a performance standard might add that this test
would involve using a keyboard. The business about eighty words a
minute would be a proficiency standard.
One of the DPIs ground rules was that it would formulate only content
and performance standards. Proficiency standards were to be left to
local school boards.
Outside the Thoughtworld, of course, content standards and
performance standards are not standards at all. A content standard is
what regular people call an academic subject; a performance
standard is a kind of assignment or activity. An actual example of a

Mulroy

content standard from the final version of the fourth grade language
arts standards is reading a wide range of materials; a related
performance standard is reading aloud. Only if you got down to
defining what or how quickly or accurately students must read would you
have a proficiency (or real) standard, but only local school boards deal
with those.
It takes a while to internalize the boldness of this sham, especially in the
face of all those golden signs announcing that Higher Standards Start
Here. If the DPI had responded to the call for higher academic standards
in normal English, they would have said, We refuse to set any statewide
academic standards. We will, however, compile lists of subjects and
activities in which local school boards may or may not impose some
standards. But Longer Lists of Subjects Start Here doesnt make a very
good motto.
Obviously, my goal of including the ability to identify parts of speech and
diagram sentences was doomed from the start. That would involve
proficiency standards. The closest the committee would come were
grammar-related performance standards, which generously
interpreted would be passed by anyone who could (say) order a
hamburger in English, e.g., (fourth grade students shall) understand and
use parts of speech effectively, including nouns, pronouns, and
adjectives. Nor was this result unusual. Many states are adopting new sets
of academic standards. Wisconsins are typical.2
Frustrating as this was, there was worse to come. Aware of my interest in
grammar, a DPI consultant thoughtfully gave me a printout from the
Internet on the subject of teaching grammar. The web site from which it
was taken is produced by the National Council of Teachers of English

Mulroy

(NCTE), the professional association of English teachers with a membership

of eighty thousand. The handout read as follows:
FACTS ON THE TEACHING OF GRAMMAR
Research over a period of nearly 90 years has consistently shown that the
teaching of school grammar has little or no effect on students. George
Hillocks and Michael Smith, 1991
Background
The most common reason for teaching grammar as a system for analyzing
and labeling sentences has been to accomplish some practical aim or
aims, typically the improvement of writing. For decades, however,
research has demonstrated that the teaching of grammar rarely
accomplishes such practical goals....
And so on.3
The printout contained bibliographical references so that the reader
could consult the research on which these surprising opinions are based.
To say that it is not compelling is a bit of an understatement. For example,
the linchpin of the 1991 study by Hillocks and Smith the one quoted on
the web pages marquee is a deliberately confusing test on identifying
parts of speech, which was given to Scottish high school students in 1947.4
Their poor performance is said to prove that typical students are unable to
learn how to identify the eight parts of speech.
The demise of college programs that (like Classics) make serious demands
on students verbal abilities is not the only sign of problems in K-12
language arts curriculum. Mean verbal scores on the SATs declined by 42
points between 1967 and 1993 (when scores were recentered) three
times the decline in quantitative scores. At UWM, which is not unusual in

Mulroy

this respect, a quarter of incoming freshmen are required to enroll in noncredit remedial writing courses in which they study the construction of
sentences and paragraphs. Within living memory, public school sixth
graders were required to explicate selections in McGuffeys readers,
which included unaltered scenes from Shakespearean plays and poems
like William Cullen Bryants Thanatopsis (So live, that when thy summons
comes to join/The innumerable caravan, which moves/To that mysterious
realm, where each shall take/His chamber in the silent halls of
death,/Thou go not, like the quarry slave at night,/ Scourged to his
dungeon ....)5 Flashing forward to 1998, Wisconsins DPI proposes a high
school graduation test. In the English Language Arts portion, it plans to ask
graduating seniors questions like: What can be assumed about caring for
the garment that has this care label attached? 100% cotton/Made in
USA/Machine Wash Cold/Tumble Dry Low/One Size Fits All. The correct
answer? D. This garment should not be washed in hot water.6 But now it
looks like the test wont be given after all. Too hard.
And the nations largest official organization of English teachers opposes
instruction in English grammar. Thanks to their efforts, grammar has been
banished from grammar school. Is it possible that there is some subtle
connection among these facts? To be fair, the arguments against
instruction in grammar are not utterly preposterous at first glance. Studies
do show that the addition of instruction in grammar for a year or two does
not dramatically benefit students working on English composition or
foreign language.7 From this it is inferred, however, that knowledge of
grammar is of no benefit in the mastery of those subjects, which is quite a
leap. In fact, advocates of instruction in grammar view it as a
foundational discipline best taught early in grade school. A good
foundation in grammar enables students to excel subsequently in
composition, foreign language study, and other verbal subjects.

Mulroy

Subsequently is the critical term. Obviously, students who are struggling

to write well or to learn a foreign language will not respond favorably to
being saddled with the whole new subject of English grammar. By
analogy, the history of western art clearly establishes that an
understanding of anatomy is beneficial to painters and sculptors. Still,
students in a class on figure drawing would not necessarily benefit from
the addition of an anatomy textbook to their syllabus. The class would be
less enjoyable, the students would become self-conscious and probably
perform less well than a rival class that did not bother with anatomy. The
lesson, however, is not to eliminate anatomy from the artists training, but
to provide its own appropriate place in the general curriculum. And that is
also the real lesson of studies purporting to show that grammar just
confuses students.
This is exactly the approach taken in the Brookfield Academy, one of the
few schools in the nation to fight back in the war against grammar. There
traditional grammar with a heavy emphasis on sentence diagraming lies
at the heart of the language arts curriculum in the second through the
fifth grade. Despite the NCTEs years of research, it is pretty clear that
this approach does not damage the Brookfield students. Their verbal SAT
scores were 86 points above the national average in 1999. Their teachers
also report that the students are generally enthusiastic about studying
grammar. I have visited their classes and find this to be obviously true.
One fifth grade class enjoys a game in which they are allowed to realize
abstract diagrams with gross sentences. When I challenged them to
produce a gross linking sentence whose subject was a gerund with a
direct object, every hand in the class shot up. A typical response:
Spewing chunks is unpleasant.

Mulroy

In his third great philosophical treatise, Critique of Judgment, Kant makes

an interesting distinction between determinant and reflective
judgments. A determinant judgment is one in which a set of rules or
concepts is stipulated and applied to a particular situation. For example,
a meteorologist classifying clouds as cirrus or cumulonimbus, etc. is
making determinant judgments. Reflective judgments may reach some of
the same conclusions, but they move in the opposite direction. You are
given a particular situation and asked, in effect, what you think about it.
You can apply whatever concepts come to mind. Contemplating clouds,
you could say that they are pretty or threatening, look like cotton, are
rapidly approaching from the west, promise rain or even that they are
cirrus clouds. The difference is that in reflective judgments the choice of
concepts is open.
Reflective judgments are relaxed because they let the mind act freely.
They are rarely wrong because the person who makes them uses
whatever rules or concepts come to mind and judges only those details of
a situation that he notices. Determinant judgments are much harder. You
lose your freedom. You have to play by given rules, understand them, and
remember them correctly. If you dont notice all the relevant details, you
will be wrong.
Classifying words by part of speech, parsing verbs, and diagraming
sentences are all examples of the use of determinant judgments and the
criticisms against them as pedagogical assignments can be made
against all intellectual activities that rely primarily on determinant
judgments. And all the criticisms boil down to one: they are hard.
For years progressive educators have done what they could to minimize
the use of determinant judgment in education. The war against grammar

Mulroy

10

is just one of their more successful efforts in this area. The effort to enhance
student freedom by minimizing determinant judgments is profoundly
misguided. Education serves no more important purpose than developing
the capacity to make accurate determinate judgments, which are
essential to every practical endeavor. One consults doctors, lawyers,
mechanics for determinant diagnoses, not reflective impressions. Not only
that, the exercise of purely reflective judgments, even as an academic
assignment, quickly becomes boring, precisely because it is not
challenging and because the evaluation of reflective judgment is
subjective. Nothing is either right or wrong.
The standard public school curriculum now provides little opportunity for
the exercise of determinant judgment in the language arts beyond the
basic skill of reading. Everything else depends on the students reflective
judgments: how they choose to express themselves, what books they like
and why, and so forth. Such a curriculum is a good formula for producing
ennui. To retain interest, academic areas need nuclei of systematic
knowledge whose use requires determinant judgments. In language arts,
this is provided by the approach formerly taken in parochial schools and
now used in places like the Brookfield Academy: first reading, then English
grammar, and then the study of foreign language with a grammatical
syllabus.
Of course, it would be wrong to go to the other extreme and emphasize
nothing but determinant judgments. What good teachers always do is to
look for activities that synthesize reflective and determinant judgments,
fostering freedom within constraints. Nothing prevents the inclusion of the
systematic study of grammar in such a mix.

Mulroy

11

To put it another way, making up any old gross sentences is boring, but
making up gross linking sentences whose subjects are gerunds with direct
objects can be a lot of fun.

Notes
1. Victor Hanson and John Heath, Who Killed Homer? The Demise of
Classical Education and the Recovery of Greek Wisdom (New York: The
Free Press, 1998) p. 3.
2. English language standards proposed or adopted by twenty-eight
states are reviewed by Sandra Stotsky, State English Standards, Fordham
Report I.1 (Thomas B. Fordham Foundation) July 1997. 3.
4. Hillocks and Smith, Grammar and Usage, Handbook of Research on
Teaching the English Language Arts (New York: MacMillan, 1991). The
Scottish test given by one W. J. Macauley gave preference to words that
could be used as different parts of speech, e.g., dance as a verb in one
sentence and a noun in another; daily as an adverb and then as an
adjective. Hillocks and Smith also rely heavily on a book published in 1952
by one Charles Fries, The Structure of English (New York: Harcourt, Brace).
This book is a shrill attack on traditional school grammar, whose basic
ideas are presented in an oversimplified way, ridiculed, and dismissed. It is
true that the concepts of traditional grammar were derived from Latin
and must be stretched somewhat to fit English and other languages. Still, if
Hillocks had consulted more scholarly works, e.g., Huddlestons
Introduction to the Grammar of English, published by the Cambridge
University Press in 1984, he would have learned that the search for
universal principles of grammars in recent decades has made scholars

Mulroy

12

more -- not less -- sympathetic with traditional grammar because

traditional grammar still provides linguists with the only starting point for
understanding the deep-seated similarities among all languages.
5. The first McGuffeys readers appeared in 1836. The series remained in
widespread use for nearly a century. It is estimated the 122 million copies
were sold between 1836 and 1920. The 1879 edition of the sixth grade
reader was the last one to contain major revisions. It has been reprinted as
a Signet Classic with an introduction by Henry Steele Commager:
McGuffeys Sixth Eclectic Reader, 1879 Edition (New York: New American
Library, 1963).
6. The sample test is no longer available on the DPI web site. For a hard
copy, contact the author or the Office of Educational Accountability,
State of Wisconsin Department of Public Instruction, PO Box 7841,
Madison, WI 53707-7841.
7. The Hillocks and Smith article cited on the NCTE web page (see above
note 4) attempts to camouflage this limitation. Of all the studies they have
reviewed, they say, by far the most impressive is by Elley, et al. (1976). In
this study, students were divided into three groups, one studying
generative or transformational grammar, one studying no grammar, and
one described by Hillocks and Smith as studying TSG (traditional school
grammar). Whereas other studies are acknowledged to have been done
on too short-term a basis to be persuasive, Elley and associates consider
the achievement of New Zealand high school students as they moved
through the third, fourth, and fifth forms and in a follow-up one year after
the completion of the instruction. After three years, there were no
significant differences among the groups. The conclusion: teaching
grammar does not have a beneficial effect on students writing. If you

Mulroy

13

consult the study, however, you find that its main purpose was specifically
to determine the direct effects of a study of transformational grammar
on the language growth of secondary school pupils. An incidental
observation is that students whose curriculum contained elements of
traditional grammar showed no measurable benefits. There is no
indication of how much traditional grammar was taught. The authors
mention that the transformational grammar group was given additional
tests on the central concepts of that approach. Nothing similar is
mentioned for the TSG group. In any event, this most impressive of
studies has nothing to say of the benefits of learning grammar
systematically in grade school, which is the point at issue.
David Mulroy is Associate Professor of Classics at the University of
Wisconsin-Milwaukee.