Last revised 8/21/1999
Dr. Ed Vavra's KISS Approach to Sentence Structure
Return to the Essays on Grammar Menu

Louis David

# 2: 
On Learning Those Pesky Parts of Speech
     Ken Donelson, while the editor of English Journal, obviously derided the teaching of grammar when he related a conversation on a plane:
        "(What do you wish you'd learned?)
         "Well, I couldn't tell you to this day what a noun was. Or a verb. Any of that."
         (He asked me so I told him -- a noun is the name of a person, place, or thing. A verb shows action or a state of being. Then I asked him if that helped.)
         "Really?" (Then he repeated almost exactly what I'd just said, and he did know them, those stupid, pointless, valueless definitions, he really did.) "Yeah, maybe it does. I'll try to remember."
         (I remember saying to myself, don't give me that bullshit.)"
      "Editor's Page: A Sort of Dumb Thing Happened on My Way to San Francisco." English Journal, March 1983.
Was it Franklin who said that a little knowledge is a dangerous thing? As editor of English Journal, Donelson, of course, held a very influential position. Unfortunately, his ignorance, arrogance, and simple lack of respect for others' opinions ("bullshit"?) reflected a wide-spread attitude among English teachers and especially among the members of NCTE. (See "Was NCTE Prejudiced against the Teaching of Grammar?") The situation has improved somewhat, but there is still strong antagonism, even among teachers who like grammar and want to teach it, against careful study of the parts of speech. Such antagonism is wrong -- and probably even harmful. It contributes to sloppy thinking.
      People like Donelson, who are strongly opposed to the teaching of grammar, will never be convinced. Their brains are simply turned off by the question. (Perhaps their heads are too full of -- to use Donelson's own term -- "bullshit"?) The teachers who want to teach grammar, but who are still antagonistic to the parts of speech, present a more interesting question. From what I have gathered (from fifteen years as editor of Syntax in the Schools), there are two distinct origins for their antagonism. In a sense, it is a distinction between the trenches and the towers. In the trenches, general English teachers attempt to include grammar in their teaching of English; in the towers, specialists in grammar do something else.
     In general, English teachers are very poorly prepared to teach grammar. They  usually know that what they currently teach does not work very well, but they do not know what else to teach, or how to teach it. (This is the fault of our Education Programs and of our colleges; it is NOT the fault of the teachers!) These teachers often try to teach the parts of speech -- in third grade, in fifth grade, in seventh grade, in tenth grade, in college, and in all the grades in between. Their problem is that the students don't seem to learn them, and this leads to frustration and eventually to antagonism. The students' failure to learn is, as I suggest throughout this site, the result of the way grammar is taught -- isolated bits of knowledge that are never (or at least rarely) integrated. But this failure, in turn, is the result of our entirely inadequate preparation of teachers. A few years ago, Syntax in the Schools ran a series of articles on main and subordinate clauses -- is the main idea in a sentence usually embedded in the main subject and verb?  I shouldn't have been, but I was surprised when a number of subscribers told me that the series was, for them, incomprehensible because they cannot identify clauses. And these were teachers who want to teach grammar and even belong to ATEG.
     Their situation reminded me of a state's educational standards I reviewed years ago: subordinate conjunctions were to be learned in fifth grade; subordinate clauses, in seventh. (Because I live in, and often pick on, Pennsylvania, I want to note that these were not Pennsylvania standards.) Such standards make absolutely no sense, but they are an excellent example of what I described previously -- isolated bits of knowledge that are never (or at least rarely) integrated. Subordinate conjunctions only make sense as lead-ins to subordinate clauses -- to memorize the conjunctions without understanding what they do is to make grammar meaningless. Most English teachers, given their situation, are therefore understandably frustrated -- and even antagonistic. But they can learn, and it is toward them that most of my work is directed.
     As for those in the towers, I respect them, but I'm afraid that their case is hopeless. There are, of course, far fewer people in the towers, and they are mainly professors who have devoted their entire lives to studying grammar -- various kinds of grammar. In addition to "traditional grammar," for example, there are "structural" grammars, "transformational" grammars, "systemic" grammars, "tagmemic" grammars, etc. etc. In many of these grammars, the traditional eight parts of speech have been junked and replaced with different sets of categories. (I can't remember which one, but one theory replaced the traditional eight parts with twenty six.)  It should be obvious, by now, why most members of the towers are antagonistic toward the eight parts of speech.
     Perhaps some day, some member(s) of the towers will produce an effective pedagogical grammar, but thus far I haven't seen one. [But then, I may have blinders too.] From what I have seen, however, not only grammars, but also concepts and exceptions within grammars, multiply like rabbits, but without rabbits' reasons. As a simple example, many of the new grammars distinguish "determiners" ("the," "a," "an") from normal adjectives. I can understand the theoretical reason for doing so -- within some of the theories in which this is done. But practically -- pedagogically -- speaking, what is the purpose? Some native speakers have trouble knowing when to use "a" and when to use "an." Non-native speakers have trouble with when to, and when not to, use these words. But will any of these speakers be helped by the knowledge that these words are "determiners" and not "regular adjectives"?  What I have seen thus far from the towers is multiplication of concepts, lack of agreement, and little exploration of why students should learn what the towers profess. I almost see the members of the towers as gourmet chefs, arguing about the best ingredients for a recipe, but never cooking. Meanwhile, our students are starving. Is all this fancy stuff necessary, or is there a simpler way to do it? As Louis M. Myers noted:
     The idea that grammar is a logical system has a tendency to make us concentrate on the "concepts" involved, and to turn us away from the study of the actual phenomena of language. A great deal of the material that appears in many texts leads only to the ability to talk about the language according to a set of artificial conventions, and has no value whatever is increasing our ability either to use or to understand it.
     For instance, it is necessary to distinguish between "direct" and "indirect objects" in Latin for the simple physical reason that they take different forms. In English the distinction, however fascinating, is completely useless. A four-year-old can make and understand sentences like "He gave me a book"; and he won't be able to do either a bit better for learning that  me may be called an "indirect object" and book a "direct object." We do not have separate dative and accusative cases in English; and since the boy is not in the least likely to say "He gave I a book" or "He gave a book me," no question of either form or position is involved. If we force him to "distinguish between these constructions" we are not teaching him anything about the use of language, but only about an unnecessarily complicated linguistic theory.
     It may be that all knowledge is good in itself, and that the question "Good for what?" is impertinent. But only those who choose to do so need learn about butterflies or old postage stamps; and the one justification for the widespread belief that everybody should learn something about grammar is the theory that a knowledge of it is directly useful in communication. Considering the amounts of time, money, and effort invested, we cannot afford to accept this theory in blind faith.
     As a matter of fact, a good many of our educators have stopped accepting it, and have announced that there is no discoverable connection between training in "formal grammar" and ability to use English effectively.
But instead of recommending changes in our "formal grammar," they have usually been satisfied that it should be abandoned entirely. This is throwing out the baby with the bath. English is not so different from all other subjects that a systematic study of it is worthless. We need only make sure that the system is soundly organized on the basis of the facts, and that we concentrate our attention on the choices we have to make, not on things to say when there is no question of a choice.
--- from Louis M. Myers, "Language, Logic, and Grammar," as reprinted in Thomas Clark Pollock, et. al. Explorations: Reading, Thinking, Discussion, Writing. Englewood Cliffs: Prentice-Hall, 1956, 112-113, originally from American English, a Twentieth Century Grammar, Prentice-Hall, 1952.

What we need is a simple, "soundly organized" grammar that will enable our students to perform a "systematic study" of whatever they read or write. For this purpose, the eight parts of speech are completely adequate, if not elegant.

     Even many educators who like to teach grammar miss that elegance. Why they do may be suggested by the widely read and highly respected writer about science, Stephen Jay Gould (whose passage motivated this essay). Gould writes:

     We often portray taxonomy as the dullest of all fields, as expressed in a variety of deprecatory metaphors: hanging garments on nature's coatrack; placing items into pigeonholes; or (in an image properly resented by philatelists) sticking stamps into the album of reality. All these images clip the wings of taxonomy and reduce the science of classification to the dullest task of keeping things neat and tidy. But these portrayals also reflect a cardinal fallacy: the assumption of a fully objective nature "out there" and visible in the same way to any unprejudiced observer .... If such a vision could be sustained, I suppose that a taxonomy would become the most boring of all sciences, for nature would then present a set of obvious pigeonholes, and taxonomists would search for occupants and shove them in -- an enterprise requiring diligence, perhaps, but not much creativity or imagination.
     But classifications are not passive ordering devices in a world objectively divided into obvious categories. Taxonomies are human decisions imposed upon nature -- theories about the causes of nature's order. The chronicle of historical changes in classification provides our finest insight into conceptual revolutions in human thought. Objective nature does exist, but we can converse with her only through the structure of our taxonomic systems.
     We may grant this general point, but still hold that certain fundamental categories present so little ambiguity that basic divisions must be invariant across time and culture. Not so -- not for these, or for any subjects. Categories are human impositions upon nature (though nature's factuality offers hints and suggestions in return). Consider, as an example, the "obvious" division of humans into two sexes. [Gould then goes into an interesting discussion of this question.]  
Full House: The Spread of Excellence from Plato to Darwin. New York: Harmony, 1996. 39.

Too many grammarians -- traditional and "modern" -- believe that grammatical categories are "assumption[s] of a fully objective nature 'out there' and visible in the same way to any unprejudiced observer" rather than "human impositions upon nature."  Some linguists, for example, have dismissed the KISS definition of the basic sentence pattern (subject / verb / complement) because they believe that "out there" sentences are not tri-partite, but bi-partite (subject / predicate). When pressed for the reasons for their belief, rather than pointing to and explaining actual sentences in real texts, they will usually refer to the work of other grammarians.
     But if categories are "human impositions upon nature," then before turning to other grammarians for an answer, we need to ask what reasons, in what contexts, did those grammarians have for establishing and defining their categories. Our general failure to ask these questions can often be seen in responses to grammatical questions asked on listservers such as ATEG and NCTE-Talk. Far too often, listserver responses consist of citations from grammar books -- end of response. That the responder does not explore WHY the grammar book explained it that way again reflects  the "assumption of a fully objective nature 'out there' and visible in the same way to any unprejudiced observer." Once we check the grammar books, it appears, the answer is, as Gould notes,  "obvious."
      Far too often, we who teach grammar, assume that the basic things are obvious. (But as Gould shows, they are not.) Recently, for example, I asked the following question on the ATEG listserver:

Do you consider "if" and "because" to be coordinating or subordinating conjunctions?  For example, consider the following two sentences: 

He wanted to know if it rained.
He was late because he forgot. 

Are they compound (two main clauses), or are they complex (a main clause and a subordinate clause)? 

Having always assumed that they are subordinating conjunctions (I'm as guilty as anyone else.), I was surprised by the responses that I received. The responses have been edited, and names have been omitted:

     Hmmm.  My first response, was why do the labels matter?  But then, I considered the idea that subordinating and coordinating conjunctions act differently. Although they might sound a little odd, both are clearly sentences if you switch the order. 
If it rained, he wanted to know.
Because he forgot, he was late. 
Take the word 'and' or 'or', and they aren't sentences if you switch the order. 
And he forgot, he was late.
Or he forgot, he was late. 
How could you consider either of those correct? They would need to be connecting a thought in a previous sentence.  This is not a grammatically correct use of these conjunctions, but as with many grammatically incorrect constructions, this can be used as an effective device if you know what you're doing.

     I would consider both "if" and "because" to be subordinating conjunctions. Both of the sentence seem to be complex sentences to me. 

     A colleague of mine ... has an article about the use of
"because" in classroom discourse; as I recall, the idea was that, in certain contexts, "because" did not function as a suborndinator. However, in the sentences below, both "if" and "because" look like subordinators to me. 

     "Because" is a subordinating conjunction, but "if" can have two distinct roles: 

(1) "If" is sometimes a subordinating conjunction, introducing an adverbial clause: 

        We'll have a picnic if the rain stops. 

Like other adverbials, adverbial clauses can be moved to precede the main clause: 

        If the rain stops, we'll have a picnic. 

(2) "If" can also be a complementizer (complementizing conjunction), introducing a complement clause: 

        I wonder if the rain will stop. 

Here the "if" clause acts as the direct object of the sentence.  "That" and "whether" are two other complementizers. 

     I also don't like sound of  'He wanted to know if it rained' under the reading 'if it rained' is object of 'know', not the condition under which he wanted to know something else.
I feel like I would use 'whether' or 'if it had rained or not' or
something. But these constructions are changing. Notably, the past perfect is being lost.

     I've always considered these to be subordinate (and so I was taught) - i.e. "complex." They still look that way to me now. 

      I agree that "because" is one [subordinating], but it is commonly used as a coordinating conjunction by my
students.  For example: 
1.  He never should have gone, BECAUSE if he hadn't the whole fiasco could have been avoided. 
"When" is another example: 
2.  "He had stopped in my room to check his email, WHEN I asked if I could question him on his opinions of  medical marijuana."  (student sentence) 
This is my question: what do we say about "because" here?  Is it being used as a coordinating conjunction?  Or is it a subordinating conjunction that is being used in a nonrestrictive clause rather than a restrictive one, hence the comma, as in the second sentence below? 
3.  We went for a swim BECAUSE it was hot.
 4.  We won the battle, ALTHOUGH we lost the war. 
One subordinating conjunction, while, has (according to Fowler) three functions: one temporal, one concessive, and one contrastive.  In
 5.  Wait WHILE I run up stairs. 
"while" is temporal; in 
6.  My brother lives in Manchester, WHILE my sister lives in Glasgow. (Fowler, 4th ed.) 
"while" is one of the others--both, really.  We mark the difference with the comma.  So: are there two or more kinds of subordinating conjunctions, as 5. and 6. suggest to me?  Or is it rather that subordinating conjunctions operate similarly in different kinds of clauses, as in 3. and
4.?  Or are there subordinating conjunctions that are sometimes coordinating ones, as in 1. and 2.?  Fowler, 2nd ed.., thinks so: "'while' (or less commonly 'whilst') is a conjunction of the kind called strong or subordinating, i.e. one that attaches a clause to a word or a sentence, not
a weak or coordinating conjunction that joins two things of equal grammatical value; it is comparable, that is, with 'if' and 'although', not with 'and' and 'or'."  Fowler outlines "the stages of degradation of 'while' from a strong conjunction to a weak one."  He gives the example, "White outfought Richie in every round, and the latter bled profusely, while both his eyes were nearly closed at the end."  This sort of sentence
Fowler calls "the flabbier kind of journalese."  (706) 

What do we say about these? 

Some further thoughts on the subject, if I may. 


Traditional grammar has a wordclass ('parts of speech') called 'conjunctions'. This class has two subclasses, the 'subordinating conjunctions' and the 'coordinating conjunctions'. 

This suggests that subordination and coordination are similar
types of relationships. And that the words expressing these
relationships basically do the same thing in sentences. This is,
however, not quite true. 

(a) Subordination and coordination
SUBORDINATING conjunctions become part of a new entity, which I call 'subclause'. 'Subclause' used in this way denotes a linguistic object, not the relationship of subordination. Traditionally also the wh-sentences in the following entity are called subordinate clauses: 'What you see / is / what you get' - which have a completely different structure. 

A subclause consists of two slots: the first slot is filled by
words like 'if, that, because, since, while,...', the second slot is filled by a sentence (major or minor, finite or non-finite).
     because / he didn't like her
     while   / running around in a circle
     if      / blue 

COORDINATING conjunctions sort of stay outside the linguistic objects they connect. The objects can be any type of building-block, i.e. sentences or noun groups or words or
morphemes or paragraphs, and so on. When two or more objects are connected in this way, the result can be called a 'chain', thus there are sentence chains, morpheme chains, etc. 

The connecting words are mainly these: 'and, or, but'. 


Then they embraced AND [connects two sentences]
Parrado, Canessa AND [single words] Vizintin set off up the mountain.
There was no light from the moon OR [noun groups] the stars. 

(b) Terminology
Subordinating conjunctions could be called 'subordinators', the coordinating conjunctions 'connectors'.

(c) The problem of 'when'
'When' I can't see as a subordinator, although traditionally it
is interpreted as one. 'When' is used like any other of the nine
'wh-pronouns'. While subordinators fill the first slot in a
subclause and are followed by a sentence (see above), wh-pronouns fill a functional slot in the sentence, they are sentence parts (subject, object, adverbial, part of prepositional phrases, etc).
'When' functions as adverbial (usually of time-point). 

When a wh-pronoun functions as a sentence part or as a part of a sentence part, it usually moves to the front of the sentence. But it still plays its original functional role. A sentence which has a wh-pronoun at its front, is a 'wh-sentence' (either as statement - mostly - or as question). 

The wh-statement can then be used in a great variety of ways: as subject, object, etc. When it has attributive function, we
usually call it 'defining relative clause'; if its function leans
towards being an appositive, we call it 'non-defining relative
clause'. We find wh-statements in cleft sentences. And in the
conclusion slot of paragraphs (especially which-sentences). And we find them on their own, e.g. in headlines (How the horse came to the American West).

Examples with WHEN- and WHERE-SENTENCES - nobody would call 'where' a subordinating conjunction, so why should 'when' be one? 

     As he followed he kept an eye on the path and watched to see / WHEN to make his move.
     I think I know / WHERE he is. 

     I'll be here / WHEN you need me.
     Andrew shoved Alyssa under the table as part of the steel
     frame crashed down / WHERE she stood. 

     Why don't you read up [ on / WHERE we're going/ ]
     Our reservations staff will advise [ on / WHEN this option is available / ].
     The Guard looked out over a large room [ from / WHERE he stood / ].
     So erm Yes. Oh I used to pay a year from when you [ from / WHEN you first started paying / ] 

     Still, it was obvious that the moment / WHEN the Sorcerer came back down / would be my best chance.
     When you entered my portal, you were transported to a place / WHERE creatures your people only dream about / truly exist.


When a subclause is embedded (not connected!) in another
sentence, the subclause fills some functional slot. 

(a) Subclauses as sentence parts
Subclauses may fill a sentence slot like subject, object,
adverbial, etc. Then they are part of the sentence in which they fill a slot. 

     Boon could see / that the snake was very big.
     to see / if they came to the top [if=whether]
     He did not see Boon / because he was busy. (adverbial of

(b) Subclauses as parts of sentences parts
The entity in square brackets is an appositional structure with
the that-clause as one of the two appositives. (The whole thing is a verbless sentence, in the sentence from which it can be thought to be derived it would play the role of subject

     Parrado walked back to Canessa, clutching the bread in his hands, [ a tangible sign / that they finally had made contact with the outside world/].

(c) Subclauses and textgrammar
Sentences are not the end of the grammatical ladder. Above
sentence grammar there comes text grammar. Many problems  of traditional grammar have to do with not recognizing that there are also grammatical rules beyond the sentence. 

There are a number of levels beyond the sentence level, a central one is the paragraph level. Paragraphs have functional slots in the same way that sentences have functional slots (subject, subject complement, etc.). The basic paragraph type offers 4 slots:
     topic slot
     description slot
     antithesis slot
     conclusion slot 

In sentences some slots are (structurally) necessary (dependent on the full verb, see the valence discussion some time ago), others are free, e.g. time adverbials or manner adverbials can be added freely to the kernel of a sentence when there is a communicative need. 

Similarly, in paragraphs the topic and the conclusion slots are
necessary to make a paragraphs, the other two are free slots. 

Back to subclause-containing structures. Intuitively we sometimes feel that there is a difference between structures that contain, say, conditional clauses (if-clauses) and structures that contain, say, that-clauses. To solve the problem we sometimes think this might have to do with subordination and coordination. 

I rather think that all this is a problem of textgrammar. E.g. a
structure like 

 If we ration it, it should last us for 20 days.
might be interpreted as a (kernel) paragraph: 

     TOPIC:       IF we ration it
     CONCLUSION:  it should last us for 20 day. 

Similarly, structures with so-that-clauses could perhaps be
interpreted in this way: 

     TOPIC:      There Canessa untied it and sent back the strap
     CONLCUSION: SO THAT their clothes, sticks, knapsacks and shoes could be thrown across in the same way. 

Both subclause-structures cannot be interpreted as adverbials, there is neither an adverbial of condition nor one of result. 

As to because-sentences - they are often adverbials of reason, but when there is comma before 'because', they, too, might be interpreted as kernel paragraphs.

Any comments on the above suggestions? 

I apologize to my colleagues if I have misrepresented any of their responses -- copying and pasting from listserver messages can lead to errors, especially where quotation marks and italics are concerned. My primary point here is the variety of responses received to something that I had not thought was even a question. [I asked it because, in reading a book about language development, I found the author assuming that "if" and "because" join main clauses.]
       The responses (presented in the order in which they were received) range from an initial (and then revised) reaction that labels do not matter, through some sharing of my assumption, through some challenges of my assumption, and then to a highly labeled explanation indeed. And these responses are to a simple(?) question as to whether or not "if" and "because" belong in the category subordinate conjunctions. It might be fun (or perhaps frustrating) to argue the answer to that question here, but the more important question is -- Why is the question itself -- and the answer(s) to it -- important?
        I do not want to suggest that any of the responses are not important, but echoing Myers, I want to know important for what? The last (detailed) response, parts of which I question, and parts of which I do not understand, may be very important for some linguistic questions, but what could anyone do with it in our classrooms -- classrooms where most students cannot even identify verbs? Before I invested time and energy in attempting to understand these concepts -- and the assumptions underneath them -- I want some evidence that all these terms have a practical pedagogical application. As for the simpler responses to the question -- "if" and "because" are or are not subordinate conjunctions -- the difference of opinion has MAJOR pedagogical implications. The two options lead to radically different definitions of "main clause," and (I hope there is some agreement here.) "main clause" is the fundamental concept of the grammar that we are and should be teaching in our schools.
     But I would suggest that, for the purposes of the K-12 classroom, WE SHOULD ALL AGREE ON THE DEFINITION, and, UNOBVIOUSLY, we don't.  Imagine the confusion -- and frustration -- of students who, in one classroom, are told that in "He was late because he forgot," "because he forgot" IS a subordinate clause, and, in another classroom -- perhaps in the same school, they are told that it IS NOT. Such disagreement is important at advanced levels of study. Not only does it demonstrate Gould's proposal that "categories are human impositions upon nature," it should also remind those of us in the towers that grammatical categories, unlike the ten commandments, are not set in stone. We need to constantly be ready to reevaluate our definitions -- and our reasons for making them.
     In our K-12 classrooms, on the other hand, our students need solid, not shifting, ground upon which to build a foundation. Our linguists understand that there are different theories of grammar, each offering different definitions of the same terms. But they should also understand that attempting to get seventh graders to comprehend all these differences -- or even bits and pieces from different theories -- only leads to the students' confusion. Our children are not stupid. Once they master the conscious ability to analyze sentence structure, once they can identify prepositional phrases, subjects, verbs, clauses (subordinate and main), once they can use these concepts to understand and discuss how sentence structure affects meaning and rhetorical effects, our students will be able to join the club and be able to understand -- and maybe even appreciate -- differences in grammatical definitions. Until then, however, we are only confusing our students. I hope my colleagues will agree, and not wish me in the position of Marat.

(c) 1999