In grave discussions of "the renaissance of the irrational" in our time, superstition does not figure largely as a serious challenge to reason or science. Parapsychology, UFOs, miracle cures, transcendental meditation and all the paths to instant enlightment are condemned, but superstition is merely deplored. Is it because it has an unacknowledged hold on so many of us?

This is the first paragraph and the author's introduction. His technique for catching the reader's attention is to associate his topic with various other contemporary manifestations of interest in the irrational. He then focuses on superstition with the rhetorical question suggesting that many people, including the reader, may be subject to superstitious behavior.

Few people will admit to being superstitious; it implies naivete or ignorance. But I live in the middle of a large university, and I see superstition in its four manifestations, alive and flourishing among people who are indisputably rational and learned.

The second paragraph states the writer's main idea: superstition is alive and well, even within the intellectual bastion of a university. He also prepares the reader to follow his discussion by mentioning that he will deal with four manifestations of superstition.

You did not know that superstition takes four forms? Theologians assure us that it does. First is what they call Vain Observances, such as not walking under a ladder, and that kind of thing. Yet I saw a deeply learned professor of anthropology, who had spilled some salt, throwing a pinch of it over his left shoulder; when I asked him why, he replied, with a wink, that it was "to hit the Devil in the eye." I did not question him further about his belief in the Devil: but I noticed that he did not smile until I asked him what he was doing.

This paragraph follows the conventional form for exposition. Davies makes a transition from the preceeding paragraph with the rhetorical question which repeats the idea that superstitions take four forms. The second sentence names the first of these: Vain Observances.

With the ground set, he defines Vain Observances by example and focuses on the particular practice of tossing spilled salt over one's shoulder.

In the final sentence, the writer makes his comment. The comment emphasizes his main point: Even intelligent and educated people are subject to superstitious practices and take them seriously. He wryly notes that the professor "did not smile until I asked him what he was doing."

Below you can see a dependency analysis of the paragraph. Some of the sentences have been truncated to make formatting easier.

1. You did not know that superstition takes four forms?
         2. Theologians assure us that it does.
                  3. First is what they call Vain Observances, such as not walking under a ladder
                  3. Yet I saw a deeply learned professor of anthropology,
                           4. when I asked him why, he replied, with a wink,
                                    5. I did not question him further about his belief in the Devil .
                                    5. but I noticed that he did not smile until

Dependency analysis is essentially a method for outlining the content of a text. When one sentence comments on another sentence, it is subordinate to it in the outline. When two (or more) sentences each refer back to a particular sentence, they are both subordinate to it but are coordinate with each other. Coordination is often signaled by a coordinating conjunction. A dependency analysis may not only show the relations among complete sentences, but it may also break individual sentences down into their component parts and show the relations among the clauses, phrases, and words.

However, for my purposes here, it is usually enough to show only the relations among complete sentences, although in some cases subordinate sentence elements have been broken out and represented in the outline. To read more about dependency analysis, consult my glossary and read this page.

In the analysis of Davies' paragraph shown above, the second sentence is subordinate to the first because the pronoun "it" has as its antecedent the word "superstition." The third and fourth sentences are both subordinate to the second because they give specific examples of superstitions. The coordination of the two examples is signaled by the use of "yet" to introduce the second one.

The next sentence (number 4 in the outline) is subordinate to the sentence that describes the professor's actions. The subordination is indicated by the word "why" which refers to why the man threw the salt over his shoulder.

The last two sentences give the writer's comment on what he observed, and a comment on an action is logically subordinate to a description of that action. In this example the comment is brief and constitutes a relatively small proportion of the paragraph. In many cases the opposite is true. A relatively brief example may be expounded upon at length and require more than one paragraph.

But as Davies shows, much can be conveyed in relatively few words. He not only makes his point that the educated can be subject to superstitious behavior, but he also indicates his own attitude. He finds the behaviour of the professor simultaneously amusing and perplexing.

The second form is Divination, or consulting oracles. Another learned professor I know, who would scorn to settle a problem by tossing a coin (which is a humble appeal to Fate to declare itself), told me quite seriously that he had resolved a matter related to university affairs by consulting the I Ching. And why not? There are thousands of people on this continent who appeal to the I Ching, and their general level of education seems to absolve them of superstition. Almost, but not quite. The I Ching, to the embarrassment of rationalists, often gives excellent advice.

The outline for this paragraph is given below.

1. The second form is Divination, or consulting oracles.
            2. Another learned professor I know, who would scorn to settle a problem by
                  3. And why not?
                        4. There are thousands of people on this continent who appeal
                        4. and their general level of education seems to absolve them
                                   5. Almost, but not quite.
                                   5. The I Ching, to the embarrassment of rationalists,

A good writer keeps the reader with him. Davies does this in his topic sentence by letting the reader know that he is going to discuss the second class of superstition, Divination. He defines the term as "consulting oracles" and then gives an example. His comment begins with the rhetorical question "And why not?" Davies again hedges his judgment of superstitious behavior. He acknowledges that many educated people consult the I Ching and that it often proves helpful, but he also implies that the apparent efficacy of the practice does not absolve the practitioners of superstition.

The third form is Idolatry, and universities can show plenty of that. If you have ever supervised a large examination room, you know how many jujus, lucky coins and other bringers of luck are placed on the desks of the candidates. Modest idolatry, but what else can you call it?

1. The third form is Idolatry,
            2. and universities can show plenty of that.
                  3. If you have ever supervised a large examination room,
                        4. Modest idolatry, but what else can you call it?

The first sentence illustrates how a statement may be followed by a comment in the same sentence. But in this case too, the comment is subordinate to the statement it refers to. The last sentence is a comment on the examples of idolotry he has observed. His attitude is that although such things as lucky coins and lucky pens may not be outlandish, they still constitute superstitious practice among those who would presumably be immune.

The fourth form is Improper Worship of the True God. A while ago I learned that every day, for several days, a $2 bill (in Canada we have $2 bills, regarded by some people as unlucky) had been tucked under a candlestick on the altar of a college chapel. Investigation revealed that an engineering student, worried about a girl, thought that bribery of the Diety might help. When I talked with him, he did not think he was pricing God cheap, because he could afford no more. A reasonable argument, but perhaps God was proud that week, for the scientific oracle went against him.

1. The fourth form is Improper Worship of the True God.
            2. A while ago I learned that every day, for several days, a $2 bill
                  3. Investigation revealed that an engineering student, worried about a girl,
                        4. When I talked with him, he did not think he was pricing God cheap,
                                          5. A reasonable argument, but perhaps God was proud

Davies knows the value of a topic sentence in clearly pointing the direction of the discussion, and so we have "the fourth form" of superstition named for us.

What follows is an example of Improper Worship of the True God, the bribe of a $2 bill. The example is followed by an explanation of the student's actions and the writer's comment. As in the preceding paragraphs, the last sentence expresses Davies' attitude, amusement at the superstitious behavior of those who are "indisputably rational and learned."

Superstition seems to run, a submerged river of crude religion, below the surface of human consciousness. It has done so for as long as we have any chronicle of human behaviour, and although I cannot prove it, I doubt if it is more prevalent today than it has always been. Superstition, the theologians tell us, comes from the Latin supersisto, meaning to stand in terror of the Diety. Most people keep their terror within bounds, but they cannot root it out, nor do they seem to want to do so.

1. Superstition seems to run, a submerged river of crude religion,
            2. It has done so for as long as we have any chronicle of human behaviour,
            2. and although I cannot prove it,
            2. Superstition, the theologians tell us,
                  3. Most people keep their terror within bounds,
                  3. but they cannot root it out,
                  3. nor do they seem to want to do so.

In the preceding paragraphs, Davies has named and given examples of the four forms of superstition. He has ended each paragraph with a comment. Having completed presentation of the four forms, he now devotes an entire paragraph to an explanation of the continued prevalence of superstitious behavior. This paragraph is wholly comment.

Note the outline. It shows a number of coordinate independent clauses. Following the topic sentence is a compound sentence. In each independent clause of this sentence the pronoun "it" refers back to "superstition." The third sentence repeats the word "superstition" and tells its origin. So all three of the independent clauses are coordinate with one another and subordinate to the topic sentence.

The last three independent clauses in the paragraph are also coordinate since each of them refers back to the word "terror."

The more the teaching of formal religion declines, or takes a sociological form, the less God appears to great numbers of people as a God of Love, resuming his older form of a watchful, minatory power, to be placated and cajoled. Superstition makes its appearance, apparently unbidden, very early in life, when children fear that stepping on cracks in the sidewalk will bring ill fortune. It may persist even among the greatly learned and devout, as in the case of Dr. Samuel Johnson, who felt it necessary to touch posts that he passed in the street. The psychoanalysts have their explanation, but calling a superstition a compulsion neurosis does not banish it.

This entire paragraph is a comment on superstitious practice where Davies attempts to explain its causes. But this paragraph does not have a satisfactory topic sentence. A good topic sentence makes a statement general enough to cover everything within a paragraph. That is not the case here. The first sentence speculates on the relation between the decline of formal religion and the persistence of superstition. The second asserts that superstitious behavior emerges early in life and is found even in learned and devout adults. These are related but distinct ideas.

So what is to be done? When a writer does not provide a suitable topic sentence, the reader must do so by inference. In this case the topic sentence for the previous paragraph might serve as the topic sentence here as well.

This solution provides a superordinate for the two parallel phrases in the first sentence ("the more the teaching of formal religion declines . . . the less God appears").

1. [Superstition seems to run, a submerged river of crude religion, below the surface of human consciousness.]
            2. The more the teaching of formal religion declines, or takes a sociological form,
            2. the less God appears to great numbers of people as a God of Love,
                  3. resuming his older form
            2. Superstition makes its appearance, apparently unbidden,
            2. It may persist even among the greatly learned and devout
                  3. The psychoanalysts have their explanation
                  3. but calling a superstition a compulsion neurosis

The paragraph illustrates another important aspect of writing. This is the practice of embedding. To embed means to put one thing inside another. Think of the Russian nested dolls called matryoshka where one figure is placed inside another and that is placed inside a third and so on. In a similar way, one kind of writing may placed inside another. So while an entire paragraph may be a comment on the topic of superstition, it can have embedded within it examples of superstitious behavior, and within these there may be comments on the particular examples.

In Davies' paragraph, the dark green font marks the embedded examples while the brown font indicates the comments on the examples. The dependency analysis shows how the examples are subordinate to the general idea that they illustrate and that the comments are subordinate to the examples. A nested element in a text is subordinate to the element it is inside of.

Many superstitions are so widespread and so old that they must have risen from a depth of the human mind that is indifferent to race or creed. Orthodox Jews place a charm on their door-posts; so do (or did) the Chinese. Some peoples of Middle Europe believe that when a man sneezes, his soul, for that moment, is absent from his body, and they hasten to bless him, least the should be seized by the Devil. How did the Melanesians come by the same idea? Superstition seems to have a link with some body of belief that far antedates the religions we know -- religions which have no place for such comforting little ceremonies and charities.

1. Many superstitions are so widespread and so old that they must have risen from             a depth of the human mind that is indifferent to race or creed.
            2. Orthodox Jews place a charm on their door-posts; so do (or did) the Chinese.
            2. Some peoples of Middle Europe believe that when a man sneezes,
            2. How did the Melanesians come by the same idea?
                  3. Superstition seems to have a link with some body of belief that far antedates

In this paragraph, Davies continues his general comments on superstition. The topic sentence suggests that the rise of superstitions was contemporaneous with the advent of human consciousness. He follows with examples and then comments on the examples. The examples are shown in green font and the comments in brown. The sentence "How did the Melanesians come by the same idea?" is a hybrid. It points to a similarity between the superstitious behavior of Europeans and Melanesians, making it an example, but as a rhetorical question, it is also a kind of comment. I have chosen to place it among the examples.

People who like disagreeable historical comparisons recall that when Rome was in decline, superstition proliferated wildly, and that something of the same sort is happening in our Western world today. They point to the popularity of astrology, and it is true that sober newspapers that would scorn to deal in love philters carry astrology columns and the fashion magazines count them among their most popular features. But when has astrology not been popular? No use saying science discredits it. When has the heart of man given a damn for science?

1.[People infer historical parallels regarding the popularity of superstitions.]
            2. People who like disagreeable historical comparisons recall that when Rome
            2. and that something of the same sort is happening in our Western world today.
                  3. They point to the popularity of astrology,
                  3. and it is true that sober newspapers
                        4. But when has astrology not been popular?
                                   5. No use saying science discredits it.
                                           6.When has the heart of man given a damn for science?

The paragraph continues Davies' comments on superstition, and it follows the familiar pattern: topic sentence, examples, and comments. I have supplied a superordinate in the outline in order to deal with the first sentence which is grammatically compound, meaning that it has two independent clauses. The grammatical parallelism is also congruent with the idea expressed: Some people see historical similarities between past times when superstitions were popular and with our own times when there appears to be a rise of interest. Davies concludes the paragraph with a comment on the irrationality of superstitious behavior, reiterating his main point. Superstition is not a matter of learning or intellect, but something much deeper.

Superstition in general is linked to man's yearning to know his fate, and to have some hand in deciding it. When my mother was a child, she innocently joined her Roman Catholic friends in killing spiders on July 11, until she learned that this was done to ensure heavy rain the day following, the anniversary of the Battle of Boyne, when the Orangemen would hold their parade. I knew an Italian, a good scientist, who watched every morning before leaving his house, so that the first person he met would not be a priest or a nun, as this would certainly bring bad luck.

1.Superstition in general is linked to man's yearning to know his fate,
            2. When my mother was a child, she innocently joined her Roman Catholic friends
            2. I knew an Italian, a good scientist, who watched every morning

This paragraph continues Davies' comments on superstition. It has a topic sentence and is developed with examples. But in this case, the writer does not comment on the examples.

I am not one to stand aloof from the rest of humanity in this matter, for when I was a university student, a gypsy woman with a child in her arms used to appear every year at examination time, and ask a shilling of anyone who touched the Lucky Baby; that swarthy infant cost me four shilling altogether; and I never failed an examination. Of course, I did it merely for the joke -- or so I thought then. Now, I am humbler.

1. I am not one to stand aloof from the rest of humanity in this matter,
            2. for when I was a university student, a gypsy woman with a child in her arms
                  3. that swarthy infant cost me four shilling altogether;
                  3. and I never failed an examination.
                           4. Of course, I did it merely for the joke
                           4. Now, I am humbler.

The final paragraph is the conclusion. It has a topic sentence, it presents an example, and it ends with the author's comments. Here Davies reaffirms the tone of the essay, fascination and amusement at the apparently irrational behavior of educated and intelligent people. He includes himself in the inditement, confessing in one final example his own submission to superstition.