Here's a problem, somewhat abbreviated, that I put to my students in a semantics class on a test last spring:

Several phrasal verbs in English use the same verb but a different particle, typically a function word that can also be a preposition or adverb. The core meanings associated with these function words suggest that they are opposites: e.g., up/down, in/out, on/off. In light of this you might expect that verbs using particles with opposite meaning would themselves have opposite meaning, but in fact they often don't. Consider, for example, bring up/bring down, carry off/carry on, turn in/turn out.

Their job was to explain why verbs whose component parts suggest opposite meaning in fact don't always mean opposite things. Overall, they did pretty well, barring a few fanciful speculations for which supporting data was absent. The upshot is that many phrasal verb particles, after long use, begin to take on a life of their own that departs significantly from what common sense suggests they should mean. Despite this, phrasal verb particles show patterns that are consistent across a number of verbs.

Consider up. It's the particle that occurs in more phrasal verbs than any other. Is up a favorite phrasal verb particle because it has so many meanings, or does it have so many meanings because it is a favorite phrasal verb particle? This may be a chicken-and-egg problem.  A somewhat exasperated lexicographer has this note in the entry for the adverb up in the Oxford English Dictionary (OED):

Some uncertainty attaches to the origin and development of many of these uses [of up], the variety of which is so great that the adverb comes to present a number of highly divergent and even directly opposite senses.

So you can say, for example, Police broke up the party (that is, ended it by dispersing its constituents) or you can say Let me just tie up these loose ends (that is, unite them). So, disunited in one case, united in the other. But here, surely, it is not up that is taking on opposite meanings; the root verbs themselves do that. So why up in both cases?

Because up, here quite divorced from its core literal meaning of "towards the sky", has a completive meaning. Many phrasal verbs ending in up suggest a movement or action that results in a definitive change of state, often with a tangible outcome. If you lawyer up you hire (and thus acquire) a lawyer. If you man up you exhibit courage or some other attribute (formerly absent) that makes you more stereotypically male.

But think about some verbs that can take either up or down as a particle. Is there any opposition to be found in their meaning? Consider the following sentence pairs. I've asterisked the two that I think are unacceptable or at least questionable.

Write down this phone number.
*Write up this phone number.

Write up an article.
*Write down an article.

The up here seems to attract the same completive aspect of meaning, which is why it makes sense to write up an article (in the end, you have an article), but doesn't make sense to write up a phone number (because there's not much to unpack in a phone number, and you already had it in the first place).

Does down in a literal sense (i.e., "towards earth") play a part in write down a phone number? You might say commonsensically that the flow of ink is usually downwards in the process of writing, and that works: until you're writing with a pencil, on a piece of paper held against a ceiling, when you could still write down a phone number. Whether down has any literal role in write down is difficult to analyze because its overwhelmingly popular meaning — record in writing, produce a written record of — tends to push out our ability to look at the verb with fresh eyes. It makes sense to write down a phone number because the things we write down are largely already existing, and we simply create a written representation of them: a name, a number, measurements, thoughts. It doesn't feel quite right to write down an article because an article is something that comes into being only as you write it, and thus you normally wouldn't say that you write it down.

Here are some sentence pairs with two phrasal verbs that are similar rather than opposite in meaning, but clearly distinct in regard to how you can use them.

Don't tear up that piece of paper, I need it.
*Don't tear down that piece of paper, I need it.

They finally tore down the old Windsor Hotel.
*They finally tore up the old Windsor Hotel.

Here the completive aspect of up prevails — with things that are small and typically manageable in the hands: you tear something up and it no longer exists in integrated form. Tear down, on the other hand, seems to require an object big enough to have exhibited a literal up aspect (such as a building or a piece of machinery) before it is deprived of its integrity. And even though the teardown of a building certainly has a completive aspect, speakers prefer to emphasize the literal up/down contrast when talking about it.

To conclude, and I hope will make this rather murky situation less opaque, let's consider this pair of completely acceptable and roughly synonymous sentences.

You have to slow down just before you reach the ranch entrance.
You have to slow up just before you reach the ranch entrance.

The up in this slow up example can be thought of as partaking in the completive sense: you reach a state of slowness that didn't exist before. So why is slow down many times more frequent than slow up? What is the semantic aspect of down in slow down that makes it much more intuitive to speakers than the completive sense of slow up?

As in the case of up, we find a notation in the OED suggesting that down can also become something of a lexicographic conundrum:

Down is used with many verbs, as bring, come, get, go, put, etc.: see these verbs. In most of these the basic uses of down correspond to those given [in the numbered senses] below, while (as with other phrasal verbs) the further developments take a more idiomatic turn.

The OED entry for down (adverb) then goes on to mention its use in phrasal verbs several dozens of times, along with the idiomatic turns that it takes.  Down is common in verbs of tracking or pursuing (track down, hunt down); of fixing in place (tie/tack/nail down); of dismissing or subduing the object of the verb (shout/boo/put down); and many others. The down in slow down seems closest to the OED sense "to or at a lower number, quantity, price, or degree." That, at any rate, is where the OED exemplifies slow down.

This turn in meaning is perhaps less idiomatic than some others, because most languages have the conceptual metaphors UP = MORE, DOWN = LESS. So in our sentence above, down shares intuitive semantic space with other verbs such as wind down, shut down, close down, settle down, all of which share a sense of going from a state of greater to lesser movement or activity.

Is it safe to say that speakers gravitate towards using phrasal verb particles that share the most semantic space in their lexicon, without regard to what literal meanings those particles might have? That seems like a good working hypothesis to me. I might check it out and then check in with you later.

Click here to read more articles from Language Lounge.

Orin Hargraves is an independent lexicographer and contributor to numerous dictionaries published in the US, the UK, and Europe. He is also the author of Mighty Fine Words and Smashing Expressions (Oxford), the definitive guide to British and American differences, and Slang Rules! (Merriam-Webster), a practical guide for English learners. In addition to writing the Language Lounge column, Orin also writes for the Macmillan Dictionary Blog. Click here to visit his website. Click here to read more articles by Orin Hargraves.