At first glance, it would seem to be easy to count the number of prepositional phrases in a passage and the number of words in those prepositional phrases. In most cases, this is, in fact, true, but once one begins analyzing passages, a few problems begin to appear. In this section, we will look at some of them. Before doing so, however, I want to note that I believe that my general statistical "results" are valid -- for the purposes for which I use them.

Prepositional Phrases per Main Clause

     At this point in my research, I do not see the number of prepositional phrases per main clause as particularly significant. Such phrases, however, are crucial to syntactic development because they often act as "hooks" to which other constructions are connected. For example, in the sentence

the subordinate clause connects to the object in the prepositional phrase. If there were no prepositional phrase, there would be no subordinate clause. The number of prepositional phrases per main clause appears to be highly influenced by the number of other constructions. In the example, there are two prepositional phrases per main clause, but the second exists only because of the subordinate clause. Perhaps a more fruitful way of exploring the number of prepositional phrases would be to divide the total by the total number of clauses. Even this, however, would be suspect, because prepositional phrases also connect to appositives, verbals, etc.

% of Words in Prepositional Phrases

     On average, approximately 33% of the words in a text are in prepositional phrases. This means that if students have as their objective the ability to explain how all the words in a passage are syntactically connected, then once they master prepositional phrases, they will be one-third of the way toward their goal. And they themselves will see this. In analyzing their own texts, some students will find less than 33% of the words in prepositional phrases. I don't believe that this will de-motivate them, especially if the teacher points out that prepositional phrases usually add a lot of details. Thus the question will grow from one of simple statistical analysis of sentence structure to the larger one of detailed writing. In the analyzed texts, and in my example, prepositional phrases are placed in { }.

Verbal Tags

     Phrases such as "look at" raise a question -- is "at" part of the verb, or is it a preposition? When working with students, I will accept either explanation if the verb plus tag can be replaced with one word (such as "watch"). In counting constructions, however, a decision had to be made. Because, in the KISS approach, students find prepositional phrases before they start working on subjects and verbs, most students will have already opted for the prepositional phrase. For that reason, I have tried to remain consistent with the same option here. This has the effect of increasing the number of prepositional phrases, the number of words in them, and the percentage of words "explained" by them.

Subordinate Clauses in Prepositional Phrases

     1. My previous research suggested that subordinate clauses account for only one in 200 objects of prepositions. 2. The computer program that I use to analyze and count constructions is not set up to count prepositional phrases within prepositional phrases. In view of the preceding two items, as a general rule, prepositional phrases that have subordinate clauses as their objects were not counted. For example (s95pr14):

In this case, "from what I have seen working in a restraunt," did not count as a prepositional phrase. I feel comfortable with this because, when students are first learning to identify such phrases, I do not expect them to get the phrases with clauses as objects correct. By eliminating such phrases from the count, the count more accurately reflects what I am looking for -- the percentage of words that students will be able to "connect" simply by learning to identify prepositional phrases.

Gerunds in Prepositional Phrases

     When it comes to counting words, gerunds raise a question because of their objects. I have tried to be consistent at including the complement of the gerund in the phrase. For example (s95pr01):

My reason for including the complement is simply that, when asking the question "by what?", most students will view "putting his butt," and not "putting" as the logical answer. This answer also coincides with our psycholinguistic model, which suggests that the bonds between an S / V / and C are stronger than the bonds of modification.