If a particle belongs to a noun connected with a preposition, the particle will, as often as not, come between the preposition and the noun (this would be absolutely impossible in Russian). Here are a few examples of this use: The younger, Leander, was above all young, it seemed to him, charmingly, crashingly so, with only a slightly greater than usual grace... (BUECHNER)

The position of these parts of the sentence is probably more free than that of all other parts. Thus, a direct address can come in almost anywhere in the sentence, as will be seen from the following few examples: "Child, I'll try." "Oh, bat, Dotty, we can't go." "Look here, Renny, why don't you come and work for me?" "Her smelling salts, Scarlett!" "What does that mean, Mr Kennedy?" (all from M. MITCHELL) "Instantly, Lieutenant, instantly." (SHAW)

We pointed out above (see p. 238) that the position of the attribute as a part of the sentence is not certain. In this section we assume that it is a part of the sentence, and treat it accordingly.

The position of adverbial modifiers in the sentence is known to be comparatively more free than that of other parts. However, there is some difference here between types of modifiers. Those which are most closely linked with the part of the sentence they modify are the ones that denote the frequency or the property of an action. They come between the subject and the predicate, or even inside the predicate if it consists of two words — an auxiliary and a notional verb, or two elements of a compound predicate.

The term "inversion" has sometimes been used to denote an unusual position of a secondary part of the sentence, that is, of an object or an adverbial modifier. That, however, is undesirable, since it might lead to misunderstandings and seriously hamper the study of word order. To illustrate our point, let us compare the following two sentences: This he knew very well, and, A pretty paradise did we build for ourselves. (THACKERAY, quoted by Poutsma) In both sentences the object stands at the beginning, which is not its usual place. After this, in the first sentence, come the subject and the predicate in their normal order for a declarative sentence, whereas in the second sentence the predicate comes before the subject. It is natural to say that in the first sentence there is no inversion, while in the second sentence there is one.

The term "word order" is a singularly unhappy one, as it is based on a confusion of two distinct levels of language structure: the level of phrases and that of the sentence. To approach this problem from a viewpoint doing justice to modern linguistic theory, we should carefully distinguish between two sets of phenomena: the order of words within a phrase and the order of parts of the sentence within a sentence. Here we are again confronted with the problem of the attribute: if the attribute is a secondary part of the sentence, its place falls under the heading "order of the parts of the sentence"; if, on the other hand, the attribute is part, not of a sentence, but of a phrase, its place with reference to its head word must be considered within the theory of the phrase and its parts. Since this question has not been settled yet, we may consider the place of the attribute in this chapter.

As we pointed out above (p. 235), the term "loose" was first used in English grammatical theory with reference to appositions.

Loose adverbial modifiers are perhaps more frequent even than loose attributes. This is especially true of those adverbial modifiers which do not modify any particular part of the sentence but refer to the sentence as a whole. They are often found at the beginning of the sentence and they point out the place, time, or the general conditions in which the action takes place.

These may be expressed by the same kind of words and phrases as the usual attributes. Their peculiarity is, that they are separated from their head word by a pause, by an intonation of their own, and by a punctuation mark (usually a comma) in writing. In actual speech such loose attributes often acquire additional shades of meaning, for example, causal or concessive, which are not expressed by any specific means, lexical or grammatical, and entirely depend on the meanings of the words in the sentence.

The theory of loose parts of the sentence is another backward element of syntactic theory. Even the terminology in this field is far from certain. The term "loose" is used in English grammars chiefly with reference to the apposition: close apposition and loose apposition are two notions opposed to each other in grammatical theory. Another term which may be used is "detached": detached attributes, detached adverbial modifiers, and so forth. We will use the term "loose".