Critical Concepts

Category Violation

and Category Errors



Consider the following three strings of English words:

(Example 1)  Old the whistling boys loudly and tunes are.

(Example 2)  The boys are loudly whisling ancient tunes.

(Example 3)  The tunes are loudly whistling ancient boys.

Only Example 2 makes sense in the language.  But the other two are nonsensical in quite different ways. 

Strings like Example 1 have sometimes been called "word salad" because it's as if a perfectly good sentence like Example 2 had been sliced to pieces and then tossed together willy nilly.  The form or recognized sense of the various words puts constraints on the grammatical roles they can play in an English sentence.   The -ly ending of loudly, for example, tells us that we have to do with an adverb, and this is a part of speech that can modify a verb, an adjective, or another adverb.  But here it doesn't occur in a place within the overall word-order of the whole string that would fit with its playing any of these roles.  Tunes might be either a plural noun (as it is in Examples 2 and 3) or a third-person singular verb (as in "He tunes pianos for a living").  Boys, on the other hand, can only be a plural noun, since English doesn't have a verb "to boy".  But, if we try to construe tunes as a verb, we can't find any noun suitably placed to serve as either its subject or its direct object.  ("To tune" is a transitive verb, and thus requires a direct object.  You can't just go around "tuning"; you have to tune something:  a flute, an engine, a football play.)  And the word-order of the string blocks us as well when we try to assign either tunes or boys some role eligible to be played by a noun  -- as a sentence subject, or direct or indirect object of a verb or verbal, or object of a preposition.  Similar problems prevent us from assigning the role of a direct article to the (the only role this word can have in English), or the role of a conjunction to and (but that is all that the English word and can do), or the role of either a participle (verbal adjective) or a gerund (verbal noun) to whistling.  The phrase the whistling boys could make sense  -- but iself cannot be assigned any sensical role within the larger word-string as a whole:  iit itself cannot be modified by ancient, and it get supplied with a verb for which it might be a subject.  Our "English machine" can't process this string into an output that is some sensical English sentence, however bizarre or false.  The string is stands to a recognizable English sentence as a well supplied construction site  -- or the scene after a tornado  -- stands to a house.  It is a chaos of raw materials waiting to be assembled  -- or trashed.

But in Example 3 something altogether different is wrong, and logicians and grammarians have coined the term "category violation" to refer to the kind of illogical or ungrammatical combinations of words or phrases we are confronted with in it.

What distinguishes Example 3 from Example 1 is precisely what it shares with Example 2:  the parts of speech that characterize each of the words in the string conform to grammatical roles permitted by the order of words in the string under one of the sentence patterns recognized in English.  On the level of the basic sentence pattern, phrase types, and parts of speech, and the particular word strings themselves, we have:

Sentence Pattern:


transitive verb

direct object

Phrase Type:

noun phrase

verb phrase

noun phrase

Parts of Speech:

article + noun  

auxiliary + adv + present participle 

adjective + noun

String (Example 1):

String (Example 2):

 The boys

 The tunes

are loudly whistling 

are loudly whistling 

ancient tunes

ancient boys

In other words, in contrast to Example 1, Examples 2 & 3 are "covered" by the rules of English grammar were taught in high school.  The problem is:  Example 3 shows that these rules are not enough.  Evidently there are essential constraints operating at a far "deeper" level than are those captured in the phrase structure and sentence structure rules governing acceptable word orders in English.  For a sentence to make sense, more kinds of things have to "fit" than what is insisted upon in the top three lines of the table we've just made.  What sorts of things are these?  First, let's get some more examples on the table.  After we're through analyzing them, we can return to the Examples 2 & 3, and you can test what you've learned by trying to analyze the difference between them.

Here are some famous artificial examples, dreamt up for purposes of illustration, and therefore deliberately flagrant. The first is by the philosopher Bertrand Russell.  The second is by the linguistic theorist Noam Chomsky.

(Example BR)  Quadruplicity drinks procrastination.

ANALYSIS:   This is an example of what is sometimes called a ‘pseudo-sentence."  It may at first glance appear to be a sentence, at least in light of the criteria we may have been taught in secondary school for recognizing whether an utterance is a sentence.  After all, it has a recognizable English noun in what we recognize as the subject slot, a recognizable English transitive verb in what we recognize as the verb slot, and another recognizable English noun in what we recognize as the direct object slot.  But it fails to make any sense, because (as it turns out) more is required for sense than meeting the conditions just cited.  Let’s analyze exactly what these essential conditions are that go unmet in the present case.  As we shall see, certain "category constraints" are violated.  We are confronted with two distinct instances of this.

"Drinking" is an action that can only be done, in the strictest sense, by non-aquatic animals with mouths (i.e., not by fish or amoebae or grass) or, in a somewhat extended sense, only by beings with mouths or, in a sense more extended still, by living things. Hence the verb "to drink" -- depending on the particular sense in which is being used -- requires as its agent something that has the logical property of falling into one or the other of these categories. Now as it happens, all of these categories in turn fall within the larger category of concrete things.

The term "quadruplicity" refers to the quality of being four-fold - shared by squares, rectangles, pairs of bisexual "swinging" married couples, dollar-stacks of quarters, etc. The suffix "-icity" marks this word as denoting an abstract quality. Since the set of abstract qualities and the set of concrete objects do not intersect - something is either one or the other, but cannot be both - quadruplicity falls outside the category of things that can perform the category of action known as drinking. To posit "quadruplicity" as the doer of the deed "drinking" is therefore to commit a category violation.

Moreover, drinking can only be done to objects that have the property of being ingestible. While we can (unfortunately) drink hydrochloric acid, we cannot drink molten steel - to say nothing of I-beams, concrete slabs, toast, computers, electronic spreadsheets, wonderment, understanding, or personality traits - all of which, for different reasons, fall outside the category of ingestible liquids.

Procrastination is the disposition to put off until later things that ought to be done sooner rather than later. Hence it lacks the logical traits that would make it eligible to be conceived as the object of the action of drinking. The transitive verb/object combination "drink procrastination" therefore constitutes another category error. Procrastination is the sort of thing that could be forsworn (or that a person could refuse to forswear). It can be given in to (or not given in to). But it can’t be drunk (or, for that matter, not drunk: it is equally nonsense to say, "I haven’t had any procrastination to drink today.")

(Example NC)  Colorless green ideas sleep furiously.

ANALYSIS:   The general pattern of this pseudo-sentence is familiar enough to speakers of English. It’s the same as we find in, for example: Witless old men babble unendingly or Hairless pink unicorns frolic joyfully - both perfectly sensible English sentences. The conditions met in the latter cases but violated in the Chomsky’s gem are imposed by the logical properties that attach to the constituent concepts evoked by the successive terms. The example is wittily rich in category violations. You will have no trouble in locating six. In fact, every combination in this specimen has been arranged to be nonsense.

Before reading further, see if you can see how.  Then check your intuitions against the following observations.

Something that is colorless can’t be green, green being a color.

Ideas are not the sort of thing that can be green - or any other color!

Yet ideas do not fall within the category of things that can be said to be "colorless," either. Water could be colorless (transparent), but not wishes. Fishes could be colorless (but just aren’t, as a matter of fact.) <Note 1>  Nor does it make sense to deny that ideas or wishes are colorless (as it does to deny that fishes or pond water are. They are simply not of the category of things which can logically either be or not be transparent.

When we say that "sleeping can’t be done by ideas" or "ideas can’t sleep," we mean something importantly different by "can’t" than we do when we say that "Bill can’t cook." Idea-[agent] and sleep-[act] don’t fit - just as, in plumbing, a ¾-inch copper elbow doesn’t fit with a ½-inch threaded galvanized pipe. If you want to stick them directly together (i.e., without a suitable adapter or intermediate run) you’re going to get a leak because this sort of connection just "doesn’t work."

It’s trickier to explain our intuition that sleeping can’t be done furiously. Is it because things done furiously are done in a state of being awake?  We can dream in fury , and when we dream, we sleep. But we don’t sleep in fury, even though we can be in a fury in our sleep. It’s as if, under the logic of the terms, sleeping can "contain" fury, but fury can’t "contain" sleep. <Note 2>

Now it's time to try your own hand.  Can you explain, in a way analogous to what we went through with the examples from Russell and Chomsky, what is wrong with Example 3?  We can make this question more precise.  The constituent concepts of both strings are the same.  That is, they have the same logical properties.  Not only that, but the grammatical roles at work in both examples are the same.  How is it then that the logical properties of the constituent concepts of Example 3 run afoul of the category requirements attaching to the grammatical roles at work in it, whereas the same logical properties satisfy these same category requirements in Example 2?  Here, for your convenience, are the two strings in question:

(Example 2)  The boys are loudly whisling ancient tunes.

(Example 3)  The tunes are loudly whistling ancient boys.

Now let's examine a few additional examples.  Consider these.

(Example 4)  The boys are loudly butchering ancient tunes.

(Example 5)  The band did a fine job of playing some wonderful ancient tunes.

(Example 6)  The band did a fine job of butchering some wonderful ancient tunes.

(Example 7)  The hunter did a fine job of butchering the deer he shot.

We see we are in even deeper water than we thought!  All of these, surely, are sensical English sentences.  But Example 4 involves a category violation.  And Example 6 involves an additional one.  How can this be, if what we have said so far is sound?  A closer look will show us the way out of this bind  -- and at the same time give us further insight into the way language works, or (in other instances) fails to work.

In fact it is nonsense to speak of "butchering tunes," if we keep to the literal meaning of "butchering" as "cutting into pieces pieces of meat suitable for cooking," since the object of this action has to be in the category "animal carcass."  Not only are melodies not kinds of animal carcasses; they are far remote from being so:  animal carcases are, for instance, members of the category "objects with mass," whereas tunes fall in the category of "immaterial objects," and these more general categories are mutually disjunct:  if something is in one, it cannot be in the other.  And to the degree we recognize the concept of "butchering tunes" as internally consistent, logically, it is only because we do not confine ourselves to the literal meaning of butchering, but have conceived another meaning on the basis of, but different from it in its logical properties, and have substituted that new meaning for the "original" or "basic" or "literal" meaning.  We mean, then, that the band "made a mess" of the work of the tunes, arranging them or performing them badly.  If our literal concept of "butchering" includes "slaughtering" (in which case its object is not an animal carcass but a food animal), then we can be advancing a second figurative dimension simultaneously:  the band is "killing" the tunes:  perhaps are committing ourselves to a figuratization of the idea tune as well, according to which a melody can be "dead" or "alive," depending on whether it makes us feel "vibrant" or leaves us repelled or "cold."

In such instances, that is, we have recuperated sense from what would otherwise be nonsense by treating one of the terms in a category violation as a figure of speech.  In constructing a figure of speech we take a term which ordinarily conveys a notion or concept that applies to a set of individual instances, and use it to convey some new notion or concept, enabling the term to refer to a different set of instances.  If this new concept logically fits the category demands imposed by the structure of the sentence as a whole together with the remaining constituent concepts, then the resulting whole can indeed "make sense," and we take the sentence to be about something different than it would have been if we had not shifted the sense of the term in this way.  In the case we have been discussing, the figure of speech that enabled us to read the string as sensical is of a sort we dub metaphor.  In metaphor, the figurative concept invoked shares some elements with the customary or literal concept (but of course in other respects must differ), so that the things referred to by the term in its figurative sense resemble in some relevant respect the things or cases the term refers to by way of the concept it customarily conveys.  In  Example 4, the band is "making a mess," just as the butcher leaves a mess at the scene of the slaughter. <Note 3>

Some hint at the complexities speakers generally have no trouble negotiating will surface if we bring to mind Example 7The hunter did a fine job of butchering the deer he shot.  Butchering (literal) is in fact a skill, and therefore can be done well or badly.  It's one thing, we might say, to butcher a steer.  But it's quite another to butcher the job of butchering a steer.  In other words, Example 4 can work even when we are quite aware that buchers don't generally "make a mess" of the work of butchering itself!

And now let's look at Example 6The band did a fine job of butchering some wonderful ancient tunes.

A person can do a good job of butchering a steer.  But "butchering a tune" is always a bad job of musicianship.  How then can it make sense to combine the figurative sense of "butchering" we resorted to in Example 4 with the idea of "doing a fine job" at it, as is posited by Example 6 The band did a fine job of butchering some wonderful ancient tunes.  You will have anticipated the answer.  Contradiction is avoided by our taking the sentence as a piece of sarcasm.  That is, we took the term "did a fine job of" as a figure of speech.  But the figure of speech in this case, though, was not metaphor (which works on the basis of similarity between what's said and what's meant), but irony.  All irony  --  and there are many varieties  --  works on the basis of discrepancy or contrast between what is said and what is meant, or taken to be meant.  In sarcasm, the discrepancy is quite stark and intense:  what is meant is an opposite to what is stated. 

In cases like Example 6 (and Example 4, which it incorporates) we do indeed have a category violation.  Indeed, it is logical incompatibility we confront in a literal construal of the string which prompts us to come up with some figurative construction that it would be intelligible, and thus appropriate to consider the speaker or writer to be meaning. <Note 4>  But there is an important difference between such instances as these and formulations such as the following:

(Example 8)  Before we start looking for a landfill site, we need to diagnose the questions we are confronted with.

(Example 9)  Determinism is a man who has no free will.

(Example 10)   Often people feel inadequate to compete with the theories and opinions that are thrown at them.

Unlike Example 3, and the famous ones by Russell and Whitehead, these examples are not the sort of thing that no one would ever find himself thinking.  They are not instances contrived to illustrate a point in a glaring way.  In fact, their insidiousness lies in the fact that they are subtle enough to actually be produced by mistake, and perhaps even without the writer's ever noticing that invisible conceptual wires have gotten crossed.

Unlike the Example 6, which prompts the reader to recover sense at the level of metaphor and irony, these sentences stubbornly resisting our attempt to turn them into something meaningful and clear by resorting to some kind of figurative construction.  In the end, they strike us as inept.  These category violations are not justified by a figurative point (however banal).  Each must count as a category error.  But not every case here is equally serious.

Example 8 was probably produced, in fact, in virtue of the richness of the writer's word-hoard.  It may be that she was groping for a thought that could for all practical purposes be either "We need to diagnose the problem ..." or "We need to get an answers to to the questions ...," and simply ended up writing down this hybrid, impossible in English, by (say) thinking the one as she began the sentence and distractedly thinking the second as she finished it.  For many of us, it's more or less inevitable that we will produce these kinds of illogical enjambments on the fly.  This sort of "overprinting" of half of one familiar idiom by half of another may well be more frequent in writing than in speech (though it happens in talk as well).  Here the solution is simply alert proof-reading.  But something similar happens a lot in speech when we find ourselves groping for words when we are on less familiar ground  -- when we have a hunch about something we might want to say, but have to cast about to see if we can think of some formulation that might fill the bill.  It typically happens that we come up with a bunch, none of them exactly right, although we'd be hard-pressed to say exactly how not.  Here, too, we may end up with a mish-mash between a couple of ideas somewhere in the neighborhood of what it might be that, if we figured it out, we really would want to think.  If we notice at the time that something sounds awkward, we may just press on, or we may stumble a bit and try to get it right.  But sometimes we just don't bother to notice, or at least dismiss the signs we may register that there is something amiss.

Something of this last sort  --  something more serious  --  seems to be at work in Examples 9  <Note 5> and 10.  These leave us suspecting that the writers of them failed to notice the conceptual glitches that prevent them from coming into focus.  We are inclined to believe that the writers are simply unaware that what they say doesn't make sense, or at least that they have not bothered (perhaps because they just didn't care) to get things clear to themselves.  Here the solution can only be to learn to appreciate what is at stake, and to commit oneself to achieving genuine insights through the process of revision.  This means proceeding from the assumption that most (if not all) insights are not reached not at a stroke, but by successive approximation, as the product of a more or less extended series of reformulations.  And it means resolving to be intellectually serious  --  to insist on discovering exactly what is not right about what one has so far thought one wanted to think, and on finding out what the thought is that one has not yet in fact conceived but that, if one could think it, would be the one to commit oneself to.

A special class of category error is comprised of illogical comparisons and contrasts. <Note 6>  Consider

Example 11a:  The flexibility of his knee is greater than his ankle.  Contrast this with
Example 11b:  The flexibility of his knee is greater than that of his ankle. Just as logical, but more simple and direct would be
Example 11c:  His knee is more flexible than his ankle.
Example 12a:  Does the theory of evolution conflict with the Creation?  Contrast this with
Example 12b:  Does the theory of evolution conflict with the Biblical story of the Creation?

For more on what is at stake in category errors, and what to do about them, see the discussion of "Microscopic Revision:  spotting and revising idomatic glitches."  For some examples of category errors that show confusion over the concept of psychological repression, see the discussion of repression.

Note 1:  The kind of "can" and "can't" we're concerned with here is logical possibility or impossibility, not factual truth or falsity, or even "factual impossibility"  --  by which we really mean "extremely high improbability," the sort o f thing we are pointing to when we say, "It's impossible for Jane to learn to play Beethoven's Kreutzer Sonata in the next two weeks."

       Still, it is true that factual and logical impossibility can in some instances take a bit of effort to untangle.  Consider what we may be assuming when we say that it's impossible Jupiter will collide with Earth next month, or that you cannot (under conditions of heat and pressure that prevail in the Earth's atmosphere at sea-level) form any such ion as would be designated by the formula [H7O]-2.  What kind of impossibility we understand ourselves to be asserting depends on the conceptions we are forming in connection with the key terms.

       If we mean by "the Earth" and "Jupiter" what many people innocent of physics and astronomy perhaps today do  --  simply big balls of matter that happen to cruise round the Sun in peculiar ways  --  then we probably won't be asserting that it is logically impossible for one to crash into the other.  At most we would probably mean, if we uttered such a sentence, simply that it struck us as outlandish to suppose that such a thing would ever happen, since (for example) it hasn't happened during all of history, and since there don't seem to be any particular signs that it's getting ready to happen now.

       But if our notions of "the Earth" and "Jupiter" are themselves tightly bound up with what modern physics and astronomy tell us of the universe, the case will be different.  If our grounds for saying that certain conjunctions between Earth and Jupiter (or between hydrogen and oxygen atoms) are ruled out by the laws of mothion and of gravity (or by the electrodynamics of electron shells), then we have to do both with the internal logical coherence of certain formal theories and with the factual question of whether this theory is adequate to account for observable facts.  Within these theories, the events cited are logically impossible.  (We are silently assuming, that is, that no other forces than the ones we are currently ware of and take account of in our theories are going to enter into the picture.)  Under Newton's 3 laws of motion, together with the law of gravity, the orbits of the Earth and Jupiter around the Sun will not cross (still less at a point occupied by both bodies at the same moment) within the next several billion years, given the known masses, positions, and velocities of these bodies, and given the fact that their masses will not change appreciably for a long, long time.  Similarly, under current atomic theory, the particular conjunction of properties designated by such a formula as [H7O]-2 (interpreted in accordance with the notation in use among chemists) is ruled out.  Suppose, however, it turned out that someone were actually to produce such a complex.  Or suppose Jupiter were suddenly to start heading our way.  We would then change our mind about the factual impossibility of the occurence in question.  But as to the theory, we would simply say that it had been discomfirmed  --  i.e., that the theory itself was false, not that under the theory it was possble after all.  In fact it is precisely because of the condition that under the theory such an event would still be impossible that we would declare it, under those factual circumstances, to be false.      Return.

Note 2:  It may seem sensical to speak of someone’s "snoring furiously."  Still, "furiously"  --  applied to snoring -- strikes us as sensical in a now-standard secondary sense  --  snoring as if there were a determined will behind it.  We could expect to hear of a steam-engine "puffing furiously." (Here, in other words, we are actually leaving the literal and passing into a metaphorical sense  --  an important shift, discussed later in this essay.  Yet this sense, too, seems out of place with sleeping  (which is, after all, consistent with being angry, as some of our dreams attest).  That is, it makes no sense  --  even under the (figurative) concept of "furiously" as "in a fiercely determined manner"(!)  --  to speak of someone's "sleeping furiously."     Return.

Note 3:  It is worth observing that there is actually an equivalent, on the level of figurative language itself, to category violation.  An important species is afforded by the phenomenon of mixed metaphor.  In his famous essay "Politics and the English Language," the British writer George Orwell called attention to several particularly grotesque examples from hack propaganda (in this case communist, though instances abound from every persuasion):

(Example GO -1)  The facist octopus has sung its swansong.

(Example GO - 2)  The jackboot is thrown into the melting pot.

The first translates into something like:  "The facist attempt to absorb Western Europe and Russia has been dealt a fatal blow; Hitler's Germany will soon collapse.  The second wants to talk about the predicament reactionary German aristocracy (who affected the fashion of wearing jackboots) found or finds itself in when it no longer dominates state and society but finds itself absorbed in a democratic society.  As Orwell observes, such writing leaves us with the impression that the writer is asleep or a kind of automaton  --  doesn't really care about what he's thinking, and is grinding out the standard clichés dictated from above.  Worse, such a functionary is not merely content to follow the party line, but actually so bored with it as not even to be committed to doing a good job in making it provoke felt reflection for other people.  This sort of "second-order category violation" thus counts as "second-order category error."

       The interesting question arises:  is there a distinction called for on the figurative level that corresponds to the distinction between justified category violation and category error?  There may be.  Shakespeare, for one, has produced some quite striking mixed metaphors.  Here's one famous example, from Macbeth (Act 2, Scene 2, l. 36ff.), just after the protagonist, as host to his trusting king, has murdered his sleeping guest.

(Example WS-1) 
Methought I heard a voice cry, "Sleep no more!
Macbeth does murder sleep," the innocent sleep,
Sleep that knits up the ravelled sleave of care,
The death of each day's life, sore labour's bath,
Balm of hurt minds, great nature's second course,
Chief nourisher in life's feast.

      It must be admitted, however, that the case is not altogether analogous to that of first-order category violation recuperated in sensical figurative construction.  The parallel does extend to the fact that many readers feel these lines to be justified by the excitement of the shift among incompatible metaphors, which they perhaps regard not as an extended metaphor gone awry but as a series of strikingly different attempts to indicate the multifarious larger injurious extent of Macbeth's deadly intervention in his sovereign's sleep (which will return to plague his own).  But the parallel breaks down in this:  not everyone agrees in this justification in terms of thematically appropriate baroque splendor.  Some find this catalogue of mutually inconsistent images merely irritating, and are prone to sigh that not only Homer but the Bard himself can nod (and even when he's talking about sleep)  --  whereas everyone agrees that Example 6, above, is justified by making sense on the figurative level, or that each of the individual first-order category violations in Shakespeare's catalogue is not nonsense but metaphor.

       Here's another instance, part of what is surely one of the most famous passages in all of English literature (Hamlet, Act 3, Scene 1, l. 58f):

(Example WS-2) 
To be or not to be, that is the question:
Whether 'tis nobler in the mind to suffer
The slings and arrows of outrageous fortune,
Or to take arms against a sea of troubles,
And, by opposing, end them.

 What happens when you try to envision someone "taking up arms" against "a sea"?   "Piffle!" cry some, who (it may be) turn this into the image of a defiant hero morally triumphant over the situation that threatens to overwhelm him (as the ocean would a man who stood firm against it with sword in hand)  --  but that surprisingly bows to determined resistance after all.  Others find this line silly instead of powerful.  For them, it's an unfortunate blemish in an otherwise powerful soliloquy.  These readers wish they had found, say, "host" instead of "sea" (since it is intelligible that one take up arms against a swarm of enemy), or "raise a dike" instead of "to take arms" (against that "sea of troubles").  We need not adjudicate this dispute.  For ort purposes it is enough to note the complainers find this second-order category violation (mixed metaphor) to be a species of disappointing nonsense  --  the second-order equivalent of a category error.  And the defenders have found a way, convincing to themselves, to construe it in a fashion that redeems it on a higher level, a third-order interpretation (as it were) that recuperates a justifying sense.       Return.

Note 4:  It is not essential for there to be a category violation present for there to be figurative meaning.  Apart from the case of puns  --  as in the names chosen for the commercial products Gatorade™ and Power-ade™, there is nothing to prevent Example 7 from being used sarcastically.  There could of course be a prompting in the speaker's tone (a sneer, say, on "fine") or in a wide-eyed overblown accompanying facial expression.  But the clue could lie obscurely in the remoter context (possibly able to be registered only by a limited group of those "in the know," who's knowledge in fact may consist of something so restricted as their prior awareness of the speaker's personal attitude towards the hunter in question), so that the line could be delivered "deadpan."     Return.

Note 5:  Statements of the form [Abstract noun] is [concrete example of something falling under that noun] are sometimes called "existential sentences."  When they are just inept -- like Example 9 (Determinism is a man with no free will) -- they count as category errors.  But if they come off as wit, then (as with other figures of speech), they don't count as category errors.  Consider, as Example 9bConfidence is working the crossword puzzle in ink.     Return.

Note 6:  False contraries and illogical contrasts can make for hilarious comedy.  I priceless example comes from The Mary Tyler Moore Show of 14 February 1976.  The situation is that Mary has been working on a documentary with a local priest, who she fears may be falling in love with her and, on that account, contemplating leaving the priesthood.  The people in the office have gotten interested in this state of affairs and, at one point just after Mary has left the office to see Father X [whose name I forget] on location, the following exchange takes place between Ted, the news anchorman, and Murray, the script-writer:

Murray:  Aw, come on, Ted.  You know nothing like that is goin' to come off.
Ted:  Well, I dunno.  Love is sorta strange.
Murray:  Whaddya mean?
Ted:  It brings together such opposite people.
Murray:  Yeah, well, what's that got to do with Mary and Father X.
Ted:  Well, he's a priest; she's a producer.  He's a man; she's a woman.  He's a Catholic; she's a Christian.