Among nonnative English speakers, easily the most movable and most easily misplaced modifier is the word “only.” In any of its three roles as adjective, adverb, or conjunction, “only” can effortlessly flit from place to place, creating as many meanings as the number of positions it perches upon in the sentence. It is, in a word, the ultimate floating quantifier, either intensifying or diminishing the semantic degree of the nouns or verbs it modifies, at times neatly linking one clause to another of its kind, but in the process baffling linguists and students of the language for the last 500 years.

Consider, for instance, the different meanings “only” creates by virtue of the five positions it takes in the following sentences:

• “Only I think Jennifer belongs to this league.” (“It’s only I that think Jennifer belongs to this league.”)

 

• “I only think Jennifer belongs to this league.” (“That’s the only thought I have at the moment: that Jennifer belongs to this league.”)

• “I think only Jennifer belongs to this league.” (“This is what I think: only Jennifer belongs to this league and no one else around here.”

 

• “I think Jennifer belongs only to this league.” (“This is what I think: Jennifer belongs only to this league and to no other.”)

• “I think Jennifer belongs to this league only.” (“This is what I think: it is only to this league that Jennifer rightfully belongs.”)

 

Then, after these five adjectival or adverbial roles, consider, too, how “only” works as a conjunction:

• In the role of “but”: “You may vote anyone you like, only vote wisely.”

 

• In the role of “and yet”: “Jennifer looks lovely, only she’s already very much married.”

• In the role of “except” or “were it not that”: “I’d like to bring Jennifer to Baguio, only that she might enjoy the place so much and stay there the whole summer.”

 

Even without its role as a conjunctive, however, “only” is already capable of creating so much ambiguity and semantic mischief if we are not careful. For instance, when describing a situation where we wanted to talk to a manager but only got as far as talking to his secretary, we probably would say “I saw only his secretary” or “I only saw his secretary,” either of which would adequately convey what happened. Then take note that a rather stilted way to say it, “I saw his secretary only,” even more faithfully describes what happened. Even so, the ambiguity remains.

The situation isn’t that bad in spoken usage, where “only” can be floated more freely without creating ambiguity. This is because a stronger stress can always be given to the word that the speaker wants “only” to modify, thus clearly establishing a clear intent and semantic linkage. We can see how this speech mechanism operates in the following spoken constructions, where the stressed words are shown in all-capital letters:

“I only saw HIS SECRETARY.” (“I saw nobody else.”)

“I only SAW his secretary.” (“Yes, I did see her, but I didn’t speak to her.”)

Taking into account the pitfalls in using “only” as a floating modifier in written prose, language experts have come up with the following recommendation: to be safe, place “only” immediately before the phrase we want it to modify. This means that in the office situation we described earlier, for instance, the safest—but not necessarily the best—written construction to describe what happened is the first version: “I saw only his secretary.” With “only” coming right before the noun phrase it modifies, “his secretary,” the construction poses the least danger of ambiguity. When spoken, however, the most natural and most felicitous version is obviously this other one: “I only saw his secretary.” It is much closer to the rhythm of speech, and it will be foolhardy for us to tinker with it simply to conform to the norms for edited or more formal prose.

For sure, there will be situations when written and spoken prose will clash head-on as to where to position “only” in a sentence. When this happens, we have to take recourse to what linguists call disambiguating qualifiers, or additional statements designed to clarify our meaning and eliminate ambiguity. This was the purpose of the parenthetical statements that accompanied the five “only”-usage examples that we took up earlier.

Those statements, of course, are not real disambiguating qualifiers because they are not part and parcel of the sentences themselves. A true disambiguating qualifier is integral to the statement, and already anticipates the ambiguity created when the main statement uses “only” as a floating quantifier. A good example is this: “I think only Jennifer belongs to this league; all the others simply fall short of the stringent requirements.”
In written prose, that’s actually the surest, most elegant way of preventing statements modified by “only” from going haywire. (March 4, 2004)

-----------
From the weekly column “English Plain and Simple” by Jose A. Carillo in The Manila Times, March 4, 2004, © 2004 by the Manila Times Publishing Corp. All rights reserved.

Click here to discuss/comment

By most contemporary counts, the word “that” ranks among the 10 most often used words in the English language. For instance, the Guinness Book of World Records ranks it seventh, only behind such common words as “the,” “of,” “and,” “to,” “a,” and “in,” in that order. The American Heritage Word Frequency Book puts “that” at ninth place with “is” and “you” ahead of it. A popular list in the Web ranks “that” tenth.

Still, none of the words in the magic 10 comes near to “that” in functional versatility. “The” and “a” practically serve only as articles, “of” and “to” only as prepositions, “in” practically always as a preposition, “is” only as a verb, and “you” only as a pronoun. Alone among them, “that” can perform the roles of no less than four of the eight parts of speech: as conjunction, as pronoun in two major senses, as adjective, and as an adverb. This makes “that” definitely one of the most hardworking multi-tasking workhorses of the English language.

Precisely because it is such a busybody, “that” is also among the most misused words in English. There is, to begin with, the evidently growing notion that when used as a subordinating conjunction, “that” can always be unilaterally dropped in the interest of brevity and euphony. Then there is the misconception—very common—that “that” is the semantic equivalent of “which” and “who,” and as such may be used interchangeably with them. Such misuses of “that” often result in grammatically truncated language, creasing the foreheads of even the most attentive listeners or forcing readers to reread sentences any number of times to ferret out their meaning.

Take this sentence from a recent business story in a leading newspaper: “The stock market surged to its highest in more than four-and-a-half years buoyed by foreign buying and hopes Congress will fast-track the passage of key tax measures before the end of the year, analysts said yesterday.” As you yourself must have experienced, one has to blink a few times to understand that sentence. It is bad enough that there is no punctuation after the main clause (the one that ends with “more than four-and-a-half years”). A comma would have made it clear that the segment that begins with “buoyed by foreign buying” is a participial phrase modifying “stock market” in the main clause. Even worse is that the compound phrase “buoyed by foreign buying and hopes . . .” is not constructed in parallel (to read “buoyed by foreign buying and by hopes . . .”).

But what really makes the sentence perplexing is the semantic black hole created by a missing “that,” which could have announced—and clearly tagged—the modifying noun clause “Congress will fast-track the passage of key tax measures . . .” for what it is.
Here’s how “that”—with some help from the comma and “by”—could have made that sentence as clear as a bell: “The stock market surged to its highest in more than four-and-a-half years, buoyed by foreign buying and by hopes that Congress will fast-track the passage of key tax measures before the end of the year, analysts said yesterday.”

Unfortunately, newspaper and magazine journalists seem to be under such pressure these days to routinely knock off “that” in such constructions, often with semantically disastrous results. 

The temptation to drop “that” from subordinate clauses and phrases appears to be endemic not only in journalism but also in advertising. Look at this sentence in a current courier services advertisement in an international news magazine: “With at least 100 people involved in the show, from actors and musicians to lighting and sound technicians, Pimlott has the seemingly impossible task of ensuring everybody works together in perfect harmony. . .” Somehow, what the statement is saying manages to squeak through, but not after confounding us and forcing us to guess whether the phrase “ensuring everybody” meant “buying insurance for them” or something else. That semantic bind would not have developed, of course, had the conjunction “that” been supplied to do its tagging work: “With at least 100 people involved in the show, from actors and musicians to lighting and sound technicians, Pimlott has the seemingly impossible task of ensuring that everybody works together in perfect harmony . . .”

We can thus see that dropping off the conjunction “that” to streamline sentences is fraught with pitfalls. Still, this should not be understood as advocating that it should not be done at all; there simply are too many instances when eliminating “that” in complex sentences can make our prose read and sound much better. We will take up the rules for that in the next essay, but until those rules are understood clearly and become second nature to us, it will be safe to always put “that” where it should be and leave it well enough alone. (October 11, 2004)

When to keep or knock off ‘that’–II

We saw in the previous essay some adverse consequences of arbitrarily preventing the subordinating conjunction “that” from doing its job. In mild cases, the “that”-less sentence makes us blink once or twice before we could understand it; in particularly bad cases, it makes us blink many times over in confusion.

To better gauge the semantic damage when we excise “that” from complex sentences, let’s arbitrarily assign the following rough measures: NB, “no blink,” for hardly any damage; 1B, “one blink,” for mildly confusing; 2B, “two blinks,” for moderately confusing; 3B, “three blinks,” for confusing; and 4B, “four blinks or more,” for dangerously confusing. To get a better feel of the nuances, of course, we must always think of the bracketed “that” as absent in the specimen sentences that will follow.

Here are what most of us can probably agree on as NB sentences: “I really thought [that] they were involved in looting the treasury.” “The team is confident [that] it can make progress because the spirit of Kaizen is so deeply entrenched in the company.” “Poultry producers and hog dealers have assured Malacañang [that] there will be enough chicken and pork in the market to last the Christmas season.” The missing “that” hardly affects the semantic integrity of the three sentences. Their meanings remain clear. Some of them even read better, both silently and aloud.

Now let’s take a look at some 1B sentences: “Their dilemma is [that] payment for their services has been delayed.” “Our problem is [that] heavy equipment keeps on rolling past our small street at night.” The absence of “that” in such sentences mildly assails the eyes and ears, but their meaning is rarely misunderstood or lost. In informal writing, by the way, we can reduce a 1B-sentence discomfort by substituting a comma for “that”: “Our problem is, heavy equipment keeps on passing our small street at night.” (When such sentences are spoken, a moderate pause right after the verb can make the meaning unmistakable.)

2B sentences, on the other hand, definitely distract; the absence of “that” momentarily makes us think that they mean something else: “They know [that] all the employees who see me want a raise.” “Stocks rose yesterday as the central bank signaled [that] it wanted to fuel economic growth by not matching a hike in US interest rates.” A second reading clarifies their meaning, of course, but our reading momentum has already been irrevocably slowed down.

In 3B sentences the distraction and confusion become profound: “The Bureau of Customs said [that] it has no evidence to prove [that] refined and raw oil products were shipped in by a syndicate, which later sold them cheaply to oil companies.” If only the first “that” is knocked off (as was prudently done in the news story where this sentence came from), the sentence would remain clear and qualify for NB level. But it certainly becomes a disconcerting 3B when the second “that” is also knocked off: “The Bureau of Customs said it has no evidence to prove refined and raw oil products were shipped in by a syndicate, which later sold them cheaply to oil companies.”

4B sentences, in turn, practically disintegrate semantically when “that” is knocked off. This time we have no choice but to restore it. There are, in fact, three specific grammatical conditions in which the conjunction “that” must be absolutely retained, and they apply neatly to our 4B sentences. These conditions, as identified in Theodore Bernstein’s Dos, Don’ts & Maybes of English Usage, are:

• When a time element intervenes between the verb and the clause: “We found last week that three strategic points in our coastline were vulnerable to attack.” Take out “that” and the sentence practically becomes nonsensical.

• When the verb of the clause is delayed so long as to make us think that the clause is not there at all: “A historian revealed that an ancient code of laws attributed to a chieftain named Kalantiaw was a hoax.” Without the “that,” we are misled into thinking that it was the historian who reported the existence of the code of laws, not the one who denounced it as a fake.

• When a second “that” is needed to clarify who said or did what: “The judge said that the accused was not the aggressor in the case and that his alleged accomplices acted in self-defense.” Take out the second “that,” and we would forever be guessing if it was also the judge who said that the accomplices had acted in self-defense.

As we can see, eliminating “that” should never be a touch-and-go affair. It is an art form, one that needs a lot of practice before we can be confident of never again inadvertently ruining our prose with a missing “that.” (October 18, 2004)
-----------
From the weekly column “English Plain and Simple” by Jose A. Carillo in The Manila Times, October 11 and 18, 2004, © 2004 by the Manila Times Publishing Corp. All rights reserved.

Click here to discuss/comment