Why are phrasal verbs so difficult?
A recent book on phrasal verbs (English Phrasal Verbs in Use, Michael McCarthy and Felicity O'Dell, CUP 2004) starts with the incontrovertible observation that the two things you really need to know about these verbs are (1) what they mean, and (2) how they function grammatically. The book goes on to give helpful advice about the syntactic behaviour of phrasal verbs, but when it comes to explaining the meaning of phrasal verbs, the authors seem a little less confident. They rightly focus on the issue of what the main phrasal verb particles 'mean', making the point that:
In some phrasal verbs, the particle has a clear basic meaning [but] most particles convey a number of different senses.
They go as far as saying that 'Sometimes the concrete meaning can help you guess the abstract meaning'. But this is a long way from providing learners with a systematic account of how the meaning of a phrasal verb can be explained in terms of the meanings of its two components. The authors can hardly be blamed for this - it is notoriously difficult to perceive any pattern in the way phrasal verbs attract meanings, and the general view is that these verbs just have to be individually learned and remembered.
Why are phrasal verbs thought to be so difficult? When a learner encounters an unfamiliar phrasal verb, s/he will usually know what the base verb means and what the particle means - but put the two together and you get something completely different. Even beginners know what put means and what off means, but that won't help them to guess the various meanings of put off. Teaching material which addresses this problem typically focuses more on explaining how phrasal verbs work, rather than on why they behave in the way they do. The learner is still left with the impression that it is all very arbitrary, very random, and that there don't appear to be any obvious 'rules'.
Language and 'rules'
This strikes me as anomalous. One of the consistent themes of corpus exploration is the gradual uncovering of systems - that is, the discovery that almost every aspect of our linguistic behaviour is 'rule-governed'. (I use the word 'rule' here not in the sense of instructions and prohibitions which collectively define 'correct' language use, but rather to refer to observable patterns of usage which occur with such regularity that they can be regarded as norms.) Some of these rules are obvious. The basic tenets of English morphology, for example, tell us that nouns usually add -s or -es in their plural form: once you know this, there is no need to memorize the plural form of every individual English noun you might want to use. Knowing rules is extremely useful for the language-learner because it reduces the load on their memory. And extensive corpus study is gradually revealing that there are a lot more of these rules and a lot more 'system' in the language than we had previously suspected. With regard to pragmatics, for example, Michael Stubbs has pointed out that:
A major finding of corpus linguistics is that pragmatic meanings, including evaluative connotations, are more frequently conventionally encoded than is often realized. [emphasis mine] (Stubbs 2001. 153)
Similarly, the whole idea of 'semantic prosody' (see for example www.hltmag.co.uk/jan00/idea.htm, www.hltmag.co.uk/jul02/idea.htm) entails a recognition that the lexical choices we make are often determined by factors which we may be unaware of and unable to articulate, but which - if we look at enough corpus data - can be perceived as forming a system.
But what about phrasal verbs? It seems very unlikely that their behaviour is not amenable to systematic explanation - but it has to be admitted that if there are rules in operation here, they are pretty hard to discern. A good place to start is by looking at the common particles from which most phrasal verbs are formed. While it doesn't really make sense to think of a particle as having a 'meaning', we can at least talk about the semantic information which it contributes to the phrasal verb of which it is a component. We can get an idea of the problem from the three different uses of in invoked by Groucho Marx in this wisecrack from the film Duck Soup (1933):
You can leave in a taxi. If you can't get a taxi, you can leave in a huff. If that's too soon, you can leave in a minute and a huff.
'Conceptual metaphor' and its explanatory potential
Back in the early days of HLT, Seth Lindstromberg showed that spatial prepositions (like in, on, and under) not only have more precise literal meanings than is sometimes recognized, but that 'their physically grounded (or "literal" or "concrete") spatial meanings are used to express and nuance dozens of abstract notions, many of them absolutely central to thought'. (see www.hltmag.co.uk/may01/lind.htm, www.hltmag.co.uk/jul01/lind.htm). He shows, for example, how the use of away in verbs such as beaver away and chatter away can be traced back to its literal use in a combination like drive away - the link being the notion of 'continuing without any clear endpoint'. This gives a clue as to how the 'phrasal verb problem' might be resolved, and this principle can be extended to all the main adverbial and prepositional particles that are used to create English phrasal verbs. And once the subject of metaphor is on the agenda, we turn inevitably to the work of George Lakoff and Mark Johnson (1980), whose notion of 'conceptual metaphor' provides such powerful insights into the way that meanings develop from the literal to the figurative. Conceptual metaphor helps to explain, for example, the underlying link between expressions like 'a heated debate, 'a fiery temper', 'seething with rage', 'an inflammatory remark' and 'getting hot under the collar'. As Lakoff and Johnson have shown, these formulations (and many more of the same type) are all connected by the fact that when English-speakers talk about anger, they unconsciously invoke a metaphor in which 'being angry is like getting hot or being on fire'. It is reasonable to suppose that the same principles may also help us to unravel the mysteries of phrasal verbs and their particles.
Most of the common phrasal verb particles are - in their basic meanings - words describing positions in space: up, down, in, out, on and off all have literal uses that relate to spatial orientation. But most spatial concepts also have figurative uses, and many of these are - if not actually universal - found in a wide variety of other languages: for example, the ideas of being 'up' or 'down' are often equated metaphorically with high and low quantities, with status and power, and with a person's mood (happy or sad). If an amount goes up it becomes larger, if it goes down it becomes smaller. Similarly, people in powerful positions are thought of as being 'high up', whereas the weak and powerless are 'down at the bottom'. And if you are 'down in the dumps' this is definitely not as good as being 'up on cloud nine'. As Lakoff and Johnson demonstrate, there is a coherent system at work here and the metaphorical development of literal spatial concepts is 'rooted in [our] physical and cultural experience'(p18). For example, the link between being up and having power has a basis in our experience of the world: if two people fight and one of them is physically on top of the other, that person usually ends up having the power.
We can see here the beginnings of a fairly systematic process, in which the basic, spatial particles develop new and more abstract meanings. As these particles combine with common verbs to form phrasal verbs, the metaphorical meanings of the particles contribute to our understanding of the whole verb. Thus, when someone leaves a powerful position, we say they step down, and if a dictator is removed from office he is brought down. In each case, we can see that the choice of particle is not at all arbitrary: the use of down - with its association with loss of power - gives us the best clue to the meanings of the phrasal verb.
Some of these ideas are exploited in Macmillan's recently-published phrasal verbs dictionary (Rundell 2005). Several earlier dictionaries of phrasal verbs have tried to tackle the issue of what particles 'mean' by including a 'particle index' that shows the various semantic features associated with each of the main particles. This is fine as far as it goes, but what limits all these attempts is that they describe but do not really explain. In the end, students are still left with a disparate collection of meaning areas which simply have to be memorized. Starting from the principle that the non-literal meanings of spatial particles are not arbitrarily assigned, we have aimed in this new book to identify the 'metaphorical chaining' (a phrase used by Lindstromberg in talking about prepositions) that may explain how a basic literal meaning can ramify to form a range of new, more abstract senses. For each of the 12 common particles that get this treatment, we have analyzed thousands of corpus instances, and the results are presented in diagrammatic form (showing the 'literal-to-figurative' process and the and the links between the various meanings), backed up by plenty of examples of phrasal verbs that exhibit each of the 'meanings' we have identified.
Case studies: some uses of over and back
To give a better idea of how this works, I'd like to look at a couple of particles and some of the meanings they have acquired. First, over: one of its basic meanings is the idea of 'moving across a space from one side to the other' (I saw one of my friends and went over to say hello to her). Linked closely to this are the (still physical) notions of passing, from one side to the other, above the top of something (A plane flew over the building, Someone ran over our dog), and the related idea of one thing covering something else (All the windows were boarded over). We can then see several further ideas that develop as metaphorical alternations on these spatial concepts, including:
- being in a higher position (figuratively) than someone else and having authority over them (Judge Davis will preside over the hearing)
- powerfully affecting someone's emotions (A feeling of desperation swept over me)
- trying to hide a truth or problem (The report skates over some vital issues)
By a similar process, we can trace the paths by which figurative notions attach themselves to another of over's literal meanings, the sense of something going over the edge of a container and flowing away. A small imaginative leap takes us from here to the idea of people being 'filled to overflowing' with emotion (Tensions finally boiled over and fighting broke out), and to the notion of something going beyond a limit (The meeting ran over by 30 minutes).
One of the senses of back is the idea of returning to a place you were in before (She left the room and came back ten minutes later). This is a basic, spatial concept. But the more abstract notion of time is often conceptualized in terms of space: for example, we talk about past events being 'behind us', and we think of the future as being 'ahead'. So returning to a former position links metaphorically to returning to an earlier time: In the story, an elderly lady looks back to her childhood before the war. This in turn takes us to the ideas of returning to a previous point in a discussion (I'd like to go back to a point you made in your introduction), or to a previous condition (He soon bounced back after his illness).
For 12 major particles (which - between them - appear in over 80% of all the phrasal verbs in this new dictionary), we have traced the process by which spatial concepts gradually attract networks of metaphorical meanings. It would be rash (and dishonest) to claim that we have thereby solved the mystery of phrasal verbs. But it was a very interesting experiment, and I believe it makes a useful contribution to discovering the underlying 'rules' that might make phrasal verbs seem less intimidating.
Michael Rundell is Editor-in-Chief of the Macmillan learner's dictionaries, and a director of Lexicography Masterclass Ltd (www.lexmasterclass.com)
George Lakoff & Mark Johnson 1980. Metaphors We Live By University of Chicago Press.
Michael Rundell (Editor) 2005. Macmillan Phrasal Verbs Plus . Macmillan Education.
Michael Stubbs 2001. 'Texts, Corpora, and Problems of Interpretation'. Applied Linguistics 22/2: 149-172.