The idea of taking phrasal verbs as single units developed little by little starting from the 16th century Bullakor W. (1586), for instance, was the first to distinguish adverbs from prepositions. He also pointed out that some prepositions may also be used adverbially (an overlap of function) and that in the compounding of prepositions with verbs, whether they precede or follow them, semantic changes can occur.
Laurel J. (1688) in comparing English and Latin verbs compounds pointed out that while in Latin prepositions precede the verb, in English they follow if. But, if Harmer. J. (1991) who firstly developed the idea of verbs and particle forming a group that could not be disassociated. In his comparisons with Latin, Greek and Hebrew he identified features that are distinctively English. He was concerned with the positional variation of particles, with oppositions of prefixed and post articled verbs, with the conversion of particle into verbs and surprisingly enough for his age, with verb- particle derivatives. Thanks to the observations made by grammarians above, “the phrasal verb gradually finds its way into English grammar".
With Smith L.P. (1992) homogeneous distributional criteria were established which allowed to draw a clear dividing line between “verb-adverb” and “verb-preposition” combinations. Not only he did focus his attention on the degree of cohesion between verb and particle to identify the two categories above, but also made resort to the criterions of stress to further distinguish them.
With Bollinger, D (1971) we finally get to a general but pinpointed overview of all the problems connected with the linguistic analysis, both descriptive and historical, of phrasal verbs. Even though he failed in giving indications as to whether and, when a verb- particle combination is a compound, he laid the foundations of contemporary studies of phrasal verbs.
Authors describe phrasal verbs using different criteria due to the variety of what they feel is necessary in description or analysis, which can make analysis difficult. The examples that follow, both from authors who focus on description and those, who choose to attempt theory based explanations, will serve to illustrate the confusion. The nature of the particle is also up for debate; the particles are called adverbs, particles, or something else entirely, and may or may not be considered part of the verb. The lack of consensus shows that there is certainly something unusual about these constructions, despite their common-place occurrence in English speech and writing.
The earliest known academic publication on English phrasal verbs in English was written by van Dongen in 1919 and published in Neophilologus. He interprets the particle of the phrasal verb as an adverb: “in order to express a complete action, verbs often require an adverb, e.g. “to hold up, to cut down” (p. 322). He argues that the word-order variation is due, not to the user’s inclinations, as had been previously supposed (p. 324), but to stress and syllable weight (p. 330). A monosyllabic adverb is more likely, he states, to not appear in the phrase-final position, while a multisyllabic adverb is more inclined to be drawn to the phrase-final position because of its weight (pp. 336-338). However, the user’s need to stress a monosyllabic adverb could draw it to the sentence-final position (p. 338). Additionally, van Dongen argues that if the adverb is “loosely” connected to the verb, or has a greater semantic contribution to the following noun phrase, it may also occur in the phrase-final position (pp. 346-347).
The next important work on phrasal verbs which is usually referenced is Gorlach, Marina (2004) explanation of the placement of nominal objects and phrasal verbs, first published in 1981. He too denotes the particle as an adverb, and does not argue that the verb and particle might function as a unit. However, he refutes the earlier idea that the word order alternation is simply due to stress or the length of the noun phrase (pp. 188-191). Unusually, he appeals to the discourse function of the verb and particle, rather than syntax or prosody, for a description of the phenomenon. The following quote illustrates his argument:
The principle governing the place of the objects is neither stress nor length nor rhythm, but something quite different: the news value which the idea denoted by the object has in the sentence. Objects denoting ideas that have news value, no matter whether they are nouns or pronouns, long or short, have end-position; those that have no such value come between verb and adverb. (Gorlach, Marina, 2004, p. 189)
Byram, M’ (2002) notion of “news value” dictates that the construction Verb-Particle NP will occur when the NP is introducing a new idea into the discourse, whereas Verb NP Particle will occur when the NP is not a new topic, either to the discourse or to the speaker and hearer.
Bollinger, D., (1971) takes a more theoretical approach to phrasal verbs, but also provides a more in-depth description of the particles. Instead of agreeing with previous descriptions which simply characterize particles as adverbs, he distinguishes three classes of particles which occur to form phrasal verbs: “Adverbs, Prepositions, and Adverb-Preposition words, [which] are set up on the basis of distributional relations ascertainable among the particles; these distributional relations are determined by the range of occurrence of particular particles with reference to a definite set of positions” (p. 37).