The order of subjects and verbs

In the vast majority of cases, the central clause elements, subjects (S), predicate verbs (V), objects (O), and predicatives (P) occur in a fixed order. Thus, as in the following examples, the subject precedes the verb, which precedes the object or predicative.

(1) The entire population (S) grieved (V) the death of their leader (O).

(2) The results (S) were (V) very surprising (P).

Adverbials have much greater freedom of position, and may occur in initial position, in final position, or even in various medial positions between other clause elements, although such placement is fairly restricted in English. The following example illustrates various adverbial positions in English.

(3) At the outset the containers were carefully monitored to avoid contamination

The main rule which is relevant here is, again, that the subject precedes the predicate verb in all but a few well-defined types of clauses.

However, there are a few cases in English where the order between subject and predicate verb is inverted, that is, when the predicate verb (or one of the verbs that belong to the predicate verb) actually precedes the subject. This is called inversion. Inversion is discussed in the sections below.

A Swedish perspective: Unmarked word order (click to expand/contract)

Swedish is also an SV(O) language, that is, the normal (or unmarked) order between the major clause elements is that the subject precedes the verb, and that the verb precedes any objects and predicatives (e.g. Kalle har en båt 'Kalle has a boat'). In this respect, English and Swedish are alike.

However, there is also a very important difference between the two languages when it comes to word order. Swedish is a V2 language, but English is not. V2 stands for 'verb second'. In a V2 language, the predicate verb is always in second position in declarative main clauses, where a declarative clause is a clause that is typically used to make a statement (e.g. You are nice, instead of Are you nice).

This claim about V2 languages needs to be commented on and qualified. To begin with, "second position" does not mean 'second word', but rather 'second constituent' or 'second clause element'. So, in a V2 language, the predicate verb is the second clause element.

Moreover, when we say that the predicate verb is always in second position in a declarative main clause in a V2 language, what we actually mean is that the finite verb of the predicate verb which is in second position. Some examples might be useful.

(1) [Peter] är lärare. 'Peter is a teacher.

(2) [Peters nya fru] är lärare. 'Peter's new wife is a teacher'

(3) [Peter] har aldrig varit i London. 'Peter has never been in London'

(4) [Mina kusiner] vill inte leka med mig. 'My cousins do not want to play with me'

The fact that Swedish, but not English, is a V2 language has not been illustrated by the examples given so far. The reason that no V2-effects have been possible to observe in these main clauses is that there has not been anything but subjects before the finite verb of the verb phrase functioning as predicate verb in these examples.

However, if we put an adverbial or an object at the very beginning of a declarative sentence, we will immediately be able to spot this crucial difference between the two languages:

(5) [Förra året] fick [Paul] fast anställning. '[Last year], [Paul] got a permanent post'

(6) [Dessa fenomen] upptäckte [vi] inte förrän igår. '[These phenomena] [we] did not discover until yesterday'


A Swedish perspective: Adverbial placement (click to expand/contract)

We will not go into detail here and discuss each and every adverbial type or class. The following cases should be mentioned and remembered, though:

(a) Since English is not a V2-language, there is a position for adverbials between the subject and the finite verb, as in (1):

(1) She probably met her husband in France.

This position is often used if there is no auxiliary in the predicate verb and the adverbial is realised by a one-word adverb phrase.

(b) The prescriptive rule that we should split infinitives is stronger in English than in Swedish. A split infinitive is when there is an adverbial, for instance a negation, between the infinitive marker to and the infinitive, as in (2):

(2) I told her to immediately leave the premises.

Even though this rule is often broken in authentic English, we ought to follow the rule when the adverbial is never, not, or only, that is, we say and write (3), instead of (4), even though the corresponding (5) is completely natural and grammatical in Swedish.

(3) I told you not to leave.

(4) *I told you to not leave

(5) Jag sa till dig att inte .

(c) It is much more natural in Swedish than in English to place heavy/complex adverbials in other positions than at the very beginning or the very end of the sentence. So, please avoid placing heavy adverbials in the middle of your English sentences. 


A special case: clauses with two subjects

Most clauses have only one subject. However, there are clause types which grammarians analyse as having two subjects. Such clauses have a 'light' subject in the form of a pronoun in the usual position before the predicate verb and another 'heavier' subject after the verb. In the following examples the two types of subject are highlighted.

(4) It was suggested that the Emperor of Austria should become the German Emperor.

(5) There were no signs of violence.

As these examples show, the first subject, referred to as the preparatory subject, can be either it or there. It is the form of the second subject, referred to as the postponed subject, that determines the choice of preparatory subject. 

  • It is used as the preparatory subject when the postponed subject is a clause.

  • There is used as the preparatory subject when the postponed subject is a noun phrase.

In clauses with there as the preparatory subject it is the noun phrase functioning as the postponed subject that determines the form of the verb. Thus, when the postponed subject is a singular noun phrase, the predicate verb is also singular, and when the postponed subject is a plural noun phrase, the predicate verb is also plural.

(6) There was a sudden pause.

(7) There were several armed uprisings after the revolution.

Cases where the predicate verb precedes the subject

While the subject normally precedes the predicate verb, there are cases where the order is reversed so that the predicate verb, or part of the predicate verb, precedes the subject. The most common instance of inverted word-order is found in yes/no questions, as in the following examples.

(8) Will (v) the government (S) survive (V) the election?

(9) Can (v) the President (S) go (V) beyond the law?

These examples illustrate the fact that when the predicate verb is made up of several verbs, only the first auxiliary verb precedes the subject.

For the purpose of forming questions in this way, English has a special auxiliary, do, which is used in clauses where the predicate verb consists of a single verb. Thus, to form a yes/no question from The President sent a message to Downing Street, we insert do before the subject.

(10)The President (S) sent (V) a message to Downing Street.

(11) Did (v) the President (S) send (V) a message to Downing Street?

Advice: Avoid emphatic do (click to expand/contract)

There is another use of do, which is restricted to cases where the writer wants to emphasise the truth of a statement. This do is heavily stressed in speech and is rarely found in academic writing. This use is illustrated in the following mini-dialogue.

A: I thought I told you to report to the Dean.
B: But I did report to the Dean.

Notice also that this use of do retains the normal S-V order.


Inverted word-order in non-questions: After negative elements

The same inverted word-order as that found in yes/no questions is also used in clauses introduced by a negative or restrictive clause element. In the following example, the initial element Not only is negative and is followed by inverted word order: could congress declare...

(12) Not only could (v) Congress (S) declare (V) war but the states were forbidden to engage in it without the consent of Congress.

Further examples (with the initial negative element highlighted):

(13) Never before have human rights been so fully and completely violated.

(14) On no account must the moisture level raise above 7 or 8 per cent.

(15) Not until the end of the hour-long conversation did the President get to the point.

Advice: Difficulties and alternative expressions (click to expand/contract)

The inverted order only occurs when the whole clause is affected by the initial negation/restriction. Thus, there are cases where what appears to be an initial negative element does not trigger inverted word order.

(1) In no time at all, the two doctors had more patients than they could handle.

Here, the meaning of the sentence is affirmative rather than negative and therefore the clause retains its normal S-V order.

Since the inverted order only occurs when a negative or restrictive element is placed at the beginning of a clause, it is always possible to avoid the issue of whether inversion applies by placing the negative element elsewhere in the clause.

In some instance, there are also clause patterns that keep an element near the beginning of a clause, but where inversion does not apply. One example is the cleft construction, as illustrated in the following example:

(2) It was not until the end of the conversation that the President got to the point.


A Swedish perspective: Inversion and basic word order (click to expand/contract)

English and Swedish are both SVO languages, so SVO is the basic word order of both English and Swedish. This means that the subject generally precedes the predicate verb and that other obligatory constituents of the clause/sentence (i.e. direct and indirect objects, subject and object predicatives, and obligatory adverbials) normally follow the predicate verb. Have a look at the following examples to see what is meant by this:

(1) [This investigation] (S) shows (V) [that Burton was right] (Od).

(2) [Laura] (S) gave (V) [her supervisor] (Oi) [something to think about] (Od).

(3) [This] (S) must be (V) [their best article so far] (Ps).

(4) [These results] (S) will make (V) [us] (Od) [famous] (Po).

(5) [He] (S) put (V) [his glasses] (Od) [on the copying machine] (A).

All these five sentences can be directly translated into Swedish and then analysed in the exact same way, which shows that English and Swedish have the same basic word order, i.e. the same order between the major clause elements.

When such claims about basic word order are made, the positions available for optional adverbials are normally not taken into consideration, that is, two languages can have the same basic word order, but differ from each other to same extent when it comes to the positions in which different types of adverbials typically occur.

It is often possible to vary the word order of a sentence, i.e. we do not always make use of the basic word order. However, when some other grammatical order between the clause elements is used instead of the basic one, this is always done for a reason.

Two versions of the same sentence which only differ from each other with regard to word order can never be used completely interchangeably. Which one of them is to be used depends on the linguistic context (i.e. the surrounding clauses and sentences) and on exactly what the writer wishes to express. It also happens that the word order is varied just for the sake of variation, but such practise had better be avoided by non-native writers, partly because of the risk of giving rise to unintended readings.

One important way of varying the word order is to "move" some constituent to the beginning of the clause/sentence. If we have another look at sentences (1) to (5) above, we can see that it is always possible to move another constituent to the beginning of the sentence, even though some of the resulting sentences can only be used in fairly limited and precise contexts. The basic word order is still the unmarked one, in the sense that it is the word order that would be used if there is no reason to do otherwise. So, if we move the direct object in (1) to the initial position, we get (6).

(6) [That Burton was right] (Od) [this investigation] (S) shows (V).

This is clearly not an unmarked sentence. We need to imagine, for instance, a sentence following this one, which establishes a contrast between what the investigation shows (i.e. that Burton was right) and what it does not show.

In a similar fashion, we can move the obligatory adverbial in (5) to the initial position to get (7) or the direct object in (5) to get (8):

(7) [On the copying machine] (A) [he] (S) put (V) [his glasses] (Od).

(8) [His glasses] (Od) [he] (S) put (V) [on the copying machine] (A).

In (7) we probably contrast

 (9) on the copying machine

with some other location where he put something else. In (8) we probably contrast the putting of his glasses on the copying machine with something else that he did not put on the copying machine (but either put elsewhere, or did not put anywhere at all). We have to be aware in all these cases of the fact that we can do very much with intonation and stress, so not everything depends on word order alone. 

OK, so now we know that Swedish and English have the same basic word order, since the English sentences (1) to (5) can be translated directly into Swedish, with the exact same word order, as in (10) to (14):

(10) Den här undersökningen visar att Burton hade rätt.

(11) Laura gav sin handledare något att tänka på.

(12) Detta måste vara deras bästa artikel hittills.

(13) Dessa resultat kommer att göra oss berömda.

(14) Han lade sina glasögon på kopieringsapparaten.

We also know that we do not always use the basic word order, and that alternative word orders are employed for certain reasons, i.e. we cannot normally use an alternative word order just because we feel like it; the alternative/marked word order must fit the context.

Even though English and Swedish have the same the same basic word order, there are also important word-order-related differences between the two languages, something that becomes apparent as soon as we start moving the corresponding constituents of our Swedish corresponding sentences.

If we start with

(15) Den här undersökningen visar att Burton hade rätt.

and move the direct object to the initial position, we also have to move the subject to the position after the predicate verb, that is, we get

(16) Att Burton hade rätt visar den här undersökningen.

instead of

(17) *Att Burton (2004) hade rätt den här undersökningen visar.

This is beacuse Swedish, contrary to English, is not only an SVO language, but also a V2 language.

V2 languages want to have their predicate verbs (or at least the finite verb of the verb phrase functioning as predicate verb) in second position in declarative main clauses (i.e. statements). By "second position" we should understand that it must be the second constituent, rather than, say, the second word.Word order is thus more about constituent order than about word order as such.

So, given that Swedish is a V2 language, it follows that (17) cannot be grammatical, since there are two constituents (Att Burton hade rätt and den här undersökningen) before the predicate verb visar.

Similarly, we cannot move sina glasögon or på kopieringsapparaten to the initial position, without also moving the subject to the position after the predicate verb, i.e. we get (18) instead of (19) and (20) instead of (21):

(18) Sina glasögon lade han på kopieringsapparaten.

(19) *Sina glasögon han lade på kopieringsapparaten.

(20) På kopieringsapparaten lade han sina glasögon.

(21) *På kopieringsapparaten han lade sina glasögon.

So, whenever some other constituent than the subject precedes the predicate verb in Swedish, the subject has to move to a position after the predicate verb, or at least to a position after the finite verb of the verb phrase functioning as predicate verb. There is a name for this unorthodox positioning of the subject after the (finite) verb in an SVO language. It is called inversion.

When the the whole predicate verb phrase precedes the subject, we talk about full inversion or complete inversion, and when it is just the finite verb that precedes the subject (while the rest of the predicate verb phrase follows it), we call this partial inversion.

Both English and Swedish have both types of inversion, but both complete and partial inversion are much more frequent in Swedish, because of the V2 constraint. The most important similarities and differences are illustrated in the table and through the examples below (PI = 'Partical Inversion; CI/PI = 'Complete Inversion or Partial Inversion'; NWO = 'Normal Word Order'):

English

Swedish


 PI  

 CI/PI 

 NWO 

 PI 

 CI/PI 

 NWO 

1. Initial adverbial

X

X

2. Initial clause negation/restriction

X

X

3. Initial object/predictaive

X

X

4. Yes/no question

X

X

5. wh-question (wh-word subject)

X

X

6. wh-question (wh-word not subject)

X

X

What this table is intended to show is, once again, that Swedish has inversion in many cases when English has normal word order (NWO). In such cases, Swedish has partial inversion (PI), if the predicate verb phrase consists of more than one verb (i.e. at least one auxiliary in addition to the main verb), and complete inversion (CI) if the predicate verb phrase consists of nothing but the main/lexical verb. This means that it must be the case that both partial and complete inversion is more frequent in Swedish than in English.

Partial inversion is sometimes called 'question word order' since it is the typical word order of questions in English. It is used in all yes/no questions (at least if we analyse questions such as Are you here? as cases of partial inversion, even though this is somewhat problematic).

Partial inversion is also used in all wh-questions in which the wh-word is not the subject. By wh-words we understand question words such as who, what, which, where, how, why, etc., most of which start with the letter combination wh.

To detrmine if the wh-word is the subject of the wh-question or not is not as difficult as it may sound. We know that all complete sentences (including questions) must have a subject. If there is no other constituent in the clause that functions as the subject (i.e. no other constituent that answers the question "Who/What VERB?"), the wh-word (or the phrase to which it belongs) is the subject. So, in the question

(22) Who did you meet yesterday?

you is the subject, while the wh-word is the (direct) object. This means that we should have partial inversion, which is also the case (i.e. did you meet, where the auxiliary did precedes the subject and the main verb meet follows it).

In the question (23) there is no other good candidate for the job of subject than the initial wh-word who (i.e who, or the possible referent(s) of who, is the one(s) who want(s)). Since the wh-word who is the subject, there is no inversion.

(23) Who wants to be a millionaire?

The difference between Swedish and English when it comes to wh-questions is that while English always has partial inversion in wh-questions in which the wh-word is not the subject, Swedish can have complete inversion:

(24) Vem träffade ni igår?

When such wh-questions are formed in English, we have to add a form of the dummy auxiliary do, if there is no other auxiliary available. If there is an auxiliary available in a wh-question in which the wh-word is not the subject, we make use of this auxiliary in both English and Swedish in order to create an inverted question, as in (25) and (26):

(25) Who will you meet in Denmark?

(26) Vem ska ni träffa i Danmark?

Before we have a look at a number of sentences illustrating all the cases in the table above, it should be mentioned that Swedish also has the type of complete inversion that English sometimes has after initial adverbials of location or direction, when the predicate verb is short and the subject is considerably heavier (i.e. longer, in terms of words or syllables) than the predicate verb.

English examples include (27) and (28), while (29) and (30) exemplify roughly the same phenomenon in Swedish (but remember that Swedish has inversion after all types of adverbials, so the structures are of course less exceptional in Swedish): 

(27) Off went the bus.

(28) On a hill far away stood an old rugged cross

(29) Bort flög kråkan. (instead of Kråkan flög bort)

(30) Under bordet stod en ko. (instead of En ko stod under bordet)

OK. Let us finally have a look at a number of linguistic examples that are intended to clearly illustrate the generalisations on which the table above are based (please note that fronted material is in italics, predicate verbs are in boldface, and subjects appear within square brackets):

1. After an ordinary initial adverbial, Swedish has partial or complete inversion (depending on the number of verbs in the predicate verb phrase), while English has normal word order.

(31)  Igår          köpte   [jag] en bil.

(32)  Förmodligen har  [han] redan   gått.

(33) Yesterday [I ] bought a car.

(34) Probably, [he] has already left.

 

2. After an initial clause negation/restriction, English has partial inversion, while Swedish has partial or complete inversion (depending on the number of verbs in the predicate verb phrase).

(35) Not until yesterday did [I] realise that he was my boss.

(36) Inte förrän igår          insåg     [jag] att   han var   min chef.

(37) Inte utan     orsak   hade de     klagat         på       hans uppförande.

 

3. After an initial object or predicative, English has normal word order, while Swedish has partial or complete inversion (depending on the number of verbs in the predicate verb phrase).

(38) (To) His mother [he] bought an umbrella.

(39) The handout [he] finished in no time.

(40) Ill [he] was, but he did not cancel his classes.

(41) (Till) sin mor      köpte   [han] ett paraply.

(42) Stödpapperet färdigställde [han] på nolltid.

(43) Sjuk var  [han], men han ställde      inte in      sina lektioner.  

(44) "Kalle" kallade [de]    honom inte.

 

4. In yes/no questions, English has partial inversion, while Swedish has either partial inversion or complete inversion (depending on the number of verbs in the predicate verb phrase).

(45) Will [they] go to Paris?

(46) Did [you] send him our proposal?

(47) Ska [de]    åka till Paris?

(48) Jobbar [de]    på Lunds universitet?

 

5. In wh-questions in which the wh-word is the subject, both English and Swedish has normal word order (SVO).

(49) [Who] made professor?

(50) [What] caused your problems?

(51) [Vem] blev   professor?

(52) [Vad]   orsakade era  problem?

 

6. In wh-questions in which the wh-word is not the subject, English has partial inversion, while Swedish has either partial inversion or complete inversion (depending on the number of verbs in the predicate verb phrase).

(53) What did [you] read yesterday?

(54) Why have [you] bought a new computer?

(55) Vad   läste [du]  igår?

(56) Varför har [du]  köpt en ny    dator?