Prepositions at the end of a sentence
Note: The debate should be impossible to accept. Please express interest in the comments section if interested. If you find a way to work around this, you automatically forfeit all seven points.
The full resolution is as follows: 'This house would prohibit the use of any preposition at the end of any sentence in Standard English.'
Preposition: In this debate, we will use the traditional definition of a preposition as a unit that precedes (subcategorises for) a noun phrase or pronoun in a sentence. This excludes intransitive prepositions and subordinating conjunctions, even though these are considered by many modern linguists as prepositions.
Sentence: A high-level syntactic constituent that is a) finite, b) either contains a subject or is imperative, and c) contains a verb phrase. (This is not a crosslinguistically accurate definition, but for the purpose of this debate, it will suffice.)
Standard English: The written variety of English that is used in formal and professional registers in the Anglophone, including but not limited to Britain, the US and Australia.
Basically, we are debating on whether Standard English should, prescriptively, prohibit prepositions at the ends of sentences.
1. Debate structure:
2. In general, all arguments should be posted inside the debate, but sources can also be provided in comments.
3. No trolling, semantics or kritiks.
4. If Pro disagrees with any of these rules, he must inform me before accepting the debate so that they can be modified.
My opponent has unfortunately misunderstood his position, viz. to support the prohibition of sentence-final prepositions. As such, we have agreed that the previous round will be annulled. Instead, we shall follow a new structure: R2 for contentions, R3 for rebuttals, R4 for defence and conclusion.
My case will be split into two portions. Firstly, I will outline the basis that underlies my entire argument: scientific prescriptivism. Next, I will offer a defence of preposition stranding in wh-constructions, and finally provide a defence of preposition stranding in passive sentences where an oblique object is raised as the subject.
Framework: Scientific prescriptivism
This is the attitude taken by 'scientific prescriptivism', advocated by grammarians like Bryan A Garner and linguists like Geoff Pullum (4). It is a kind of middle ground between extreme descriptivism/language relativism, in which 'anything goes', and extreme prescriptivsm, in which correctness conditions do not at all depend on what people actually produce in their speech, but is derived from some higher authority who can haphazardly impose rules like 'no split infinitives' on language users. (5)
Instead, the rules are based on rational discussion of the facts associated with language. To give an example of scientific prescriptivism, take the singular 'they'. Whether this should be prohibited is the subject of much debate. Under scientific prescriptivism, whether we should permit 'they' depends on whether its use will impose an additional cognitive load on the listener. If it will, then we should prohibit 'they'; conversely, if it produces no additional load, we should allow it.
So, why scientific prescriptivism? Well, when we standardise our language, our goals should be to facilitate communication. Language has other uses, like phatic communion and facilitating thought, but we do not intend to police such uses through standardisation, so our main focus should be facilitating communication. Imposing an additional mental load on the listener or reader will slow down the transmission of information, so the rules for standard English should prevent this from happening.
Scientific prescriptivism is the foundation on which my ensuing discussion will all be based.
C1) Preposition stranding in wh-constructions
By wh-constructions in this debate, I refer to two distinct but related syntactic phenomena in English: relative clauses and wh-questions. Both often exhibit an effect known as long-distance dependency. For the purpose of this debate, I will only consider relative and interrogative clauses in which a fronting effect occurs on an oblique object or prepositional phrase adjunct, which always contains long-distance dependencies.
To illustrate the meaning of the term, I'll give the following examples:
(1a) I've heard of this.
(The ∅ symbol simply indicates that there is no object there; I do not imply the existence of a 'null object'.)
(1b) contains a preposition at the end of a sentence, while (1c) does not. This is because in (1b), only the complement of of - that is, the relative pronoun which - was fronted, that is, put at the beginning of the phrase. In (1c), the entire prepositional phrase, of which, was fronted (a phenomenon known as pied piping), so there is no preposition at the end of the sentence. I will argue in this section that sentences with structures isomorphic to (1b) should not be prohibited in English and indeed, (1b) should be preferred over (1c) in certain cases.
SC1) 'Nobody speaks like that'
One compelling reason for choosing (1b) over (1c) in many cases is that people simply do not talk that way. Here is one example (2):
(2a) ?On what did you step?
While (2a) may sound like an acceptable sentence, it remains strange and no native speaker would talk like that. When people hear such sentences, it will also be likely that they need a moment of reflection to understand the sentence, since it does not conform to our usual methods of sentence formation. This will impede communication, and therefore should not be allowed in Standard English.
Moreover, to use sentences like (2a) will come across as pretentious and pedantic in any register (even formal speech) (2). Whenever we communicate with other using language, it is helpful to use the technique of speech accommodation (3), that is, imitating the speech style of the other interlocutor. This will improve solidarity among the speakers and ensure a healthy and relaxed atmosphere. Therefore, we ought to abandon constructions that can potentially be perceived as pretentious, such as (2a), which will cast a bad impression among the other interlocutor and increase the distance between the two.
Therefore, it is far healthier for standard English to rule (2b) acceptable than to force speakers to use (2a).
SC2) 'Everyone speaks like that'
My last argument was a negative one, arguing that the alternative is bad. I can flip the argument around and assert that preposition stranding is used by everyone. A quick search in the British National Corpus turns up 294 instances of 'about.' in newspapers and 228 in academia. This is only the tip of the iceberg - we did not count cases where 'about' precedes a question mark, or where the matrix clause is connected to another clause, leaving a comma after 'about' (e.g. 'There is nothing I know about, since I'm stupid.'). This is not an idiosyncratic feature of 'about'. We find 324 instances of 'of.' in newspapers and 230 in academia. (8)
While this does not immediately imply the preposition should be stranded - that would be an ad populum - it does show that most people do speak and write like that, even in formal situations, and by speech accommodation it would be a good idea to communicate with others in this style.
SC3) Prepositions sometimes transmit important information
Prepositions are often thought to be functional categories which carry little information or weight. This is untrue. As anyone who has taken an introductory course to linguistics knows, the preposition is one of the four main parts of speech (or five if one counts adverbs) (1, 6), and it can carry important information in the sentence. This is especially true when the verb and preposition form part of a lexicalised 'phrasal verb' like 'give up' or 'look after'. In this case, it would be helpful to put the preposition at the very end.
Pinker (2014) (7) gives the example of 'something to guard against' and 'he doesn't know what he's talking about'. Rewriting these into 'something against which to guard' and 'he doesn't know about what he's talking' is not only odd, but also packs too much of the information in the middle when they could have been better placed at the end.
C2) Preposition stranding in the passive voice
Next, I will consider preposition stranding in the passive voice where an oblique object is raised as the subject. This involves a change in subcategorisation. Usually, this means the agent of the predicate is either turned into an adjunct at the end or absent, whereas the object - in this case an oblique object - turns into the subject. Refer to the following examples for illustration (1, 6):
(3a) One can rely on him. (rely <agent, theme>)
In (3a), 'rely' subcategorises for the agent and theme, which become the subject and oblique object; in (3b), 'relied' subcategorises for the theme only. (This is a crude lexicalist account, but the transformationist view does not make different with regards to my argument.) (3c) is simply wrong; 'on he' itself is not a grammatical phrase, and a prepositional phrase cannot be used as a subject anyway.
The three sentences mean the same thing. (3c) is ungrammatical, so it's usually a choice between (3a) and (3b). One could argue that we can simply use (3a) in all situations - then a prohibition on (3b) in Standard English will be okay. But this is not the case: there are situations in which only the passive should be used, and there are situations in which only the active should be used. In the case of our example sentences, the passive wins.
Readers usually focus their attention on the subject. (7) (3a) forces the reader to give undue attention to 'one', which is a nameless dummy pronoun that can mean anyone. Clearly, this isn't a good subject to focus on.* (3b) instead puts the emphasis back on 'he', the person we can rely on. This is a much better strategy than using a dummy pronoun.
Therefore, (3b) will better facilitate communication than (3a), and hence, prepositions at the end of a sentence should not be prohibited in Standard English.
(1) Dalrymple, M. (2001). LexicalR08;Functional Grammar. John Wiley & Sons, Ltd.
*No, I'm not changing that to 'on which to focus'!
I thank my opponent for his arguments. I will now proceed to my rebuttals.
My opponent's main contention concerns the 'textbook definition' of the preposition as a word that precedes a noun. I will assume that he mean a noun phrase, not a noun - we cannot say 'on chair', but we can say 'on the chair', where 'the chair' is a noun. However, even without this slight error, his argument is unsound for three reasons I will outline below.
The first is a reductio ad absurdum:
P1) If the definition of prepositions necessitates that prepositions precede nouns in a sentence, then the definition of transitive verbs necessiatates that transitive verbs precede nouns.
P2) The definition of transitive verbs does not necessitate that they precede nouns.
C) The definition of prepositions does not necessitate that prepositions precede nouns.
P1) is a simple analogy that I'm sure my opponent will accept. The traditional definition of a transitive verb is a verb that precedes a noun phrase, and the traditional definition of a preposition is a verb that precedes an object (which is also a noun phrase). So the underlying assumption behind the protasis of P1) is that if the traditional definition says part-of-speech X should precede part-of-speech Y, then we must use part-of-speech X before part-of-speech Y in an utterance. Thus, the apodosis follows from the prodasis.
P2) is correct because of the following simple counterexample:
(1) He is a person whom I respect.
This is a correct sentence. 'Respect' is a transitive verb, yet it is not followed by a noun phrase. This is because of a long-distance dependency: The topic of the relative clause, whom, is linked to the object of respect.
C) logically follows from P1) and P2). In fact, (1b) of R2, an example of a preposition at the end of a sentence, is pretty much the prepositional analogue of (1) of this round.
My second argument is about subcategorisation. What the traditional definition implicitly suggests, in fact, is that a preposition should subcategorise for a noun phrase. This is not at all a problem for prepositions at the end of a sentence, because there is simply a long-distance dependency between the topic of the relative clause and the 'object' (complement) of the preposition.
My third argument, related to the second, is about empty categories. Many linguists believe that relative clauses are formed by movement. In other words, a wh-word is extracted from its normal position and subsequently moved to another position. So, (1b) of R2 can be characterised as being formed this way:
(2a) This is something ____ I've heard of which.
(2b) This is something which I've heard of t .
t is an empty category standing for trace. The extraction and subsequent movement of which leaves a trace in the original place. This trace can be said to be a noun phrase, so in fact, (2b) still satisfies the traditional definition of a preposition as a word used before nouns.
(1) Cook, V., & Newson, M. (2014). Chomsky's universal grammar. John Wiley & Sons.
I thank my opponent for his arguments. However, I must note that he has essentially conceded my R2 arguments, as he has not stated any objections to the premises of my framework (scientific prescriptivism), nor has he challenged any of the substantive arguments I presented.
However, he did offer a general response to my argument, and to that I must respond. He claims that the majority of grammarians today disagree with what I have written. While this is an argumentum ad populum and an appeal to authority, the premise is also false: The majority of grammarians today do not agree with prohibition. As the Chicago Manual of Style writes, 'That old prohibition is what we call a grammar superstition. You will not find it in any authoritative grammar book. Please see CMOS 5.176.' (1)
The CMOS is not alone in permitting this usage of prepositions. The New York Times' style guide also allows it; a copyeditor there once wrote, 'Sorry, but those teachers who promulgated those shibboleths were taking an extremely narrow view, one that apparently arose in Victorian times, possibly as people learned Latin in order to understand English grammar.' (2)
Fowler, one of the foremost grammarians of his time, wrote: 'Those who lay down the universal principle that final prepositions are "inelegant" are unconsciously trying to deprive the English language of a valuable idiomatic resource, which has been used freely by all our greatest writers except those whose instinct for English idiom has been overpowered by notions of correctness derived from Latin standards.' (3)
To quote a snarky remark by an Associated Press copyeditor, 'It’s something we’ll put up with.' (4)
The rule that prepositions should not be used at the end of a sentence has never been well-founded. In fact, it is an imaginary rule created by one poet to show the inferiority of another. (5) That it has subsequently been picked up by pedantic grammarians of the 19th century and spread through the Anglophone does not make it correct, reasonable or justified in any way. It is merely a myth, deeply-seated in educated circles till recent times, which is slowly being supplanted with the abandonment of misleading analogies to Latinate grammar. The application and imposition of such a rule in Standard English will spell disaster for communicative efficiency and linguistic elegance alike, as I have sufficiently demonstrated in my R2 arguments. It is about time that this hoary myth went to rest.
Finally, I thank my opponent for accepting my debate. In light of the arguments presented, I urge that voters vote Con. Thank you!
(5) Pinker, S. (2014). The sense of style: The thinking person's guide to writing in the 21st century. Penguin.
"prepositions at the end of the sentence are a threat to use of Grammar "