I once stopped for a long while over a line describing moonlight as quicksilver spilling across the ground. Not because it was good, but because it was smooth — so smooth that I didn't have to think, and my eyes simply slid past it.
Smooth is the most dangerous compliment you can pay a metaphor. A figure that goes down without friction usually means it has stopped working and is merely swapping one word for two.
The more I read, the more I think a metaphor’s real skill is not resemblance but relocation. You move a thing into another category and, by speaking of it there, let people see a relation they could not see before. If it moves, it is a metaphor; if it doesn’t move, it is a synonym.
“Is he thick as a plank” a metaphor? I don’t think so. Wood and stupidity have been tied together in the language for so long that saying it is just repetition. What has force is the other kind: instead of calling him dull, say that a sentence has to pass through three gates in his head before one gets out. That isn’t a metaphor; it is an action taken apart for you to watch — and it is both crueler and more accurate than the plank.
So a metaphor has an aim, and its aim is distance. Too close, and the reader feels nothing, because he recognised it on the spot. Too far, and he must build a bridge himself to get across — and if the bridge holds, that is his achievement, not yours. A good metaphor should make a reader pause, then nod. The pause is the distance; the nod is the landing. A pause with no nod is showing off. A nod with no pause is chatter.
Zhang Dai, writing about snow over West Lake, has a passage I still remember: sky and cloud and hill and water all one whiteness, and on the lake only “a trace of the long embankment, a dot of the pavilion at its centre, a mustard-seed of my boat, two or three grains of people in it.” Trace, dot, seed, grain — the smallest, most concrete measure words he could find. He never says the snow is vast or that people are small; he simply changes the ruler. Measuring a man in grains of rice puts the distance far enough away, and yet a man in that whiteness really is that size, so the landing is solid. That is the kind of aim I bow to whenever I recall it.
Some metaphors, on the other hand, are accurate and have long since stopped being true. “Time is a river” is not wrong; in two thousand years everyone has been able to catch it. That is exactly the trouble: if everyone can catch it, no one is seeing anything new. It is no longer a metaphor but currency — circulated so long that the pattern is worn off. Using it is convenient, not clever.
I have a clumsy rule of my own: finish a metaphor, leave it, and if you still remember it three days later it is probably on target; if you have forgotten it, it was only one word turned into two, decorative on the surface and moving nothing. It isn’t an elegant rule, but it is more use than “read more, write more,” because you can test it on the spot.
Deeper still, aim is a gauge of honesty. Whether a writer will spend one extra step on a single description shows up precisely here. Those who won’t borrow someone else’s ruler; borrow often enough and your own eye goes bad — you end up able to say only what everything is like, never what it is.
I don’t ask a metaphor to astonish. Astonishment lasts a moment; the book closes and it is gone. A good metaphor is responsible for something else: making you see from then on. The next time you meet the same thing, you will remember that ruler and measure it yourself. By then the metaphor is no longer on the page; it has grown into your eye.
Which is roughly all that aim means: it does not speak for you, it only opens one of your eyes.
卜世 撰

说些什么吧!