Between Heads and Phrases
Particles in English Phrasal Verbs
This thesis pursues two aims: first of all, it outlines the interdependency between the semantic and syntax in phrasal verb. If the particles leave their literal domain, they lose their lexical autonomy as they attach to the verbal head. Literal particles, like prepositions, are syntactically and semantically autonomous. Therefore, they may be treated as a subclass of the category preposition. This is not possible with particles in phrasal verbs which in facts are grammaticalized. This aspect seems to be neglected by the generative grammarians who concentrate on the explanation of particle movement. The questions which are dealt with are, above all, how to account for the particle movement and if the particle moves away from the verbal basis, which constituent then does assign the Case to the object noun phrase. One observes that there have developed different “schools” or direction of accounts, for example Johnson (1991) attempts to explain phrasal verbs in terms of a complex head analysis. He assumes that verb and particle form one lexical item. Kayne (1985) applies Small Clause analysis starting from the assumption that particle verbs and resultative constructions are similar. Particles are heads of a Small Clause with the nominal objects as their subjects. [...]
Table of Contents
2 A Brief Introduction to Phrasal Verbs
2.1 Phrasal Verbs: The Definition of Phrasal Verbs
2.2 Syntactic peculiarities
3 Lexical status of particles in VPCs
3.1 Current Debate and Status quo
3.2 Towards A Solution to the Categorial Problem
3.2.1 Prepositions as Intransitive Prepositions
3.2.2 Adverbial Particles as a Separate Category
4 Syntax of Verb Particle Constructions
4.1 Particle Movement in Generative Grammar
4.1.1 Johnson (1991):Particle Verbs as Complex Heads
4.2 Small Clause Analysis in Particle Verbs
4.2.1 Elenbaas (2006):Lexical Decomposition Analysis in Particle Verbs
5 Diachronic Perspective of Phrasal Verbs
5.1 Grammaticalization and its mechanisms
5.2 A Brief Aside to the History of Phrasal Verbs
5.2.1 From Inseparable Prefix to Particle
5.2.2 Word order SOV > SVO
5.3 Semantic Change in the Particles in VPCs: Literal >Abstract
5.3.1 Metaphorisation in Phrasal Verbs: Between Loss and Gain
7 Appendix: Abbreviations, Phrases and Symbols used
The treatment of particle verbs has experienced different stages. In the first half of the twentieth century, the descriptions of particle verbs have existed but they were scattered in many journal articles and mentioned only marginally in grammar books, for example Curme (1913/1914; 1931/1935), van Dongen (1919), Jespersen (1927), Roberts (1935), Erades (1961), Carstensen (1964) or Live (1965). The first compilatory monograph was provided by Kennedy in the year 1920. It belongs, together with the monographs by Bolinger (1971) or Fraser (1974/1976), to the most important works on phrasal verbs. Particle verbs have aroused a wide interest among scholars of various directions, educationalists, cognitive linguists and generative grammarians who face with the challenge to describe the verb-particle combinations with respect to their heterogeneity
Phrasal verbs, generally speaking, consist of verbs and particles which may be an adverb and a preposition to form more or less coherent unit with the verb. From the perspective of a learner of English as a foreign language (EFL), particle verbs are not only an important feature of English grammar whose degree of mastery reveals the command of lexicon and style. But, due to their heterogeneity and complexity, they are difficult to master and provide a vast potential for mistakes but they are also very complex and heterogeneous. According to Darwin and Grey (1999:65), many particle verbs have entirely idiomatic meaning although both parts of the construction seem very familiar to the learners. This idiomaticity leads to avoidance or to errors because phrasal verbs constitute “a syntactic [and semantic] oddity in the language” to include the quotation by Darwin and Gray (1999:65). This is also the reason why phrasal verbs are difficult to describe.
For historical reasons, many particles share the form and spatio-locative meaning with prepositions. However, if they leave the literal domain, they lose their lexical autonomy or, properties which make them prepositional or adverbial as it will pointed to in this thesis. In phrasal verbs, the metaphorically transferred particles “have lost [...] their directional-motional meaning” (Pelli 1976:80) and this enables them to be a part of phrasal verb and serve as markers of aspect. This is the reason why these elements resist being assigned a clearly cut lexical category, which makes the description of phrasal verbs within the X’-syntax framework difficult. X’-syntax operates on clearly cut lexical categories verbs, nouns, adjectives and prepositions where the particle in phrasal verbs has no place.
There exist two directions of solution to this problem: Arms and Becker (1969) and Emonds (1972/1979, 1985) suggests to assign the directional adverbs to the class of prepositions which may be transitive or intransitive. They have the choice to subcategorize for optional objects. This approach has found some followers. This is the basis for the analyses by Dixon (1982), Aarts (1989, 32008) and Huddleston (32005) who apply this notion to their own accounts to verb-particle constructions. Azzaro (1992:43) argues in favour of assigning of adverbial particles into the group of (in) transitive prepositions and sees them as a ”bridge between adverbs and prepositions in all the cases where the complement of a preposition is omitted [...]”. The opposite opinion is advocated by Cappelle (2004, 2005) who finds enough syntactic evidence against this claim and states that particles, in general, should be assigned to an independent category Prt which project their own PrtP.
This thesis pursues two aims: first of all, it outlines the interdependency between the semantic and syntax in phrasal verb. If the particles leave their literal domain, they lose their lexical autonomy as they attach to the verbal head. Literal particles, like prepositions, are syntactically and semantically autonomous. Therefore, they may be treated as a subclass of the category preposition. This is not possible with particles in phrasal verbs which in facts are grammaticalized. This aspect seems to be neglected by the generative grammarians who concentrate on the explanation of particle movement. The questions which are dealt with are, above all, how to account for the particle movement and if the particle moves away from the verbal basis, which constituent then does assign the Case to the object noun phrase. One observes that there have developed different “schools” or direction of accounts, for example Johnson (1991) attempts to explain phrasal verbs in terms of a complex head analysis. He assumes that verb and particle form one lexical item. Kayne (1985) applies Small Clause analysis starting from the assumption that particle verbs and resultative constructions are similar. Particles are heads of a Small Clause with the nominal objects as their subjects. Aarts (1989; 1992) supports largely Kayne’s (1985) analysis but and attempts to include semantic implications in his analysis and claims that the Small Clauses cannot be applied onto every particle verbs but only to those which show motion-result-relationship.
As a representative of the generation of young scholars, Elenbaas (2006) observes the discord concerning the approaches to particle movement in generative grammar and approaches the problem of particle movement in another way. She assumes that English particles are hybrid between heads and phrases that have the faculty to project phrases but not always do (cf. Elenbaas 2006:80). She (Elenbaas 2006:82-83) asserts that the particle projects due to the economy which requires that the heads are preferred to phrases. Particles, however, are heads by default. Particles, standing alone and projecting a phrase, take specifiers and adverbial modifiers. For her analysis she uses the lexical decomposition analysis which she maps on the split-INFL-analysis by Radford (1997, 2004).
Second aim of this paper is to risk attempt to provide some reason for the syntax-semantic interface within directional particles. It will be observed that today’s directional prepositions and adverbial particles are related in so far as they originate in phrasal adverbs, formerly separable prefixes (Hiltunen 1983) which, owing to their concrete meaning and strong stress, have survived the changes.
Thus, due to their favourable inherent semantics, adverbial particles in phrasal verbs have assumed three different forms and functions driven by the principle of split or divergence as proposed by Hopper (1991). Thereof, the various forms and functions of the adverbial particle have developed. In their spatio-directional meaning, the prepositions function as adverbs and prepositions. In their adverbial function, the directional particles have undergone second stage of divergence, namely they retain their adverbial function whereas in non-locative contexts, they undergo at least partial bleaching so as to grammaticalize as aspect markers as early as in the Middle English period.
This is the reason why particles in phrasal verbs are semi-dependent elements. However, what this chapter will not achieve is to decide on the lexical category of particles in phrasal verbs. Foskett (1991:50) observes that “[...] categories are not watertight compartments” because “[A] adverbs shade imperceptibly into prepositions and vice versa” as it is observed by Jacobson (1977:42). And this is true in this case.
This work consists of four chapters. The first part outlines the status quo concerning the syntactic and semantic peculiarities of phrasal verbs, relying on strongly on the works by Bolinger (1971) and Quirk et al. (1972;1985), Huddleston (2002:271-290) and Cappelle (2004; 2005). The reason is that the syntax and degree of semantic cohesion correlate as pointed out above. The semantic subsection points to the meaning of particles and the relation between space and time which result in aspectual use. This section is based largely on Brinton (1985; 1988/1994), the temporal analyses by Kenny (1963) and Vendler ( 1974).
The semi-dependent status of particles is the topic of the second chapter. It starts with the fundamental debate on the status of particles in general, that is how they to be placed in the realm of the parts of speech. From traditional grammar, particles are subdivided into adverbs, prepositional adverbs and prepositions. It will be shown that the traditional classification cannot be mapped onto the generative grammar where the lexical categories are strictly defined with respect to the set of distinctive features [±N; ±V], and the phrase structures which determine the number and the kind of complements and specifiers. Moreover, as mentioned above, points to the solution proposed to solve the categorial problem and the outcome is that the solution by Cappelle (2004; 2005) scores better than to assign the prepositions into the realm of prepositions although this proposal, on the first sight, is attractive.
Third chapter deals with the generative perspective of particle verbs. It is an exposition of the solutions proposed to the problematic of particle movement within transitive particle verbs. One distinguish largely among two big directions of approaches, namely complex head hypothesis as proposed by Johnson (1991), a rather popular Small –Clause analysis as originally proposed by Kayne (1985), and supported by Aarts (1989) and a fragment of the doctoral dissertation by Elenbaas (2006). It is special with respect to its assumption that particles are hybrid between heads and phrases. Elenbaas’s (2006) approach combines all the well-known from semantics, syntax and discourse which flow into her theory.
The fourth chapter deals with the development of particles in phrasal verbs. A brief excursion to the history of English has to be made. It leads to the particle status in the Old and Middle English periods where major changes have taken place. Major trigger, as it will be shown, is the reanalysis of the word order from OV > VO has caused the collapse of the morphological system. Moreover some language change processes which have contributed to the status quo or the multifunctionality of particles are very briefly sketched.
On this spot, some brief remarks have to be added. Due to the lack of space, some aspects are left out or mentioned very briefly and left without explanation. The readers are asked for indulgence towards many shortcomings of this thesis and its author in this respect.
Last but not least:to obtain a university degree, it takes a lot of patience, work and support from family and friends. The time for the expression of my gratitude to persons who have directly contributed to the genesis and development of this work.
I would like to thank Professor Gisa Rauh for leaving me creative freedom in choice and handling of the topic of the present thesis and for understanding for my situation.
I would thank my parents, Danuta and Boleslaw Haba who bore with me all the time and to whom I owe so much. They supported me materially all over the long period of my studies. Thanks for the never-ending love in times when their patience begun to wear thin. They always found the belief that the moment of my graduation will come. My brother Rafael supported me technically and he is thanked for it, too. Nicole, I love you, too. You are the sunshine in my life.
Aunt Astrid and Uncle Walter Brinkmann, thanks for your their kindness, moral support and encouragement when I needed it most. You are the best.
My teacher and friend, Thomas Prast, is thanked for having shown me the way that with the patience and a little bit of courage everything will end well.
Finally, this thesis is dedicated to the loving memory of my grandparents, Stefania Hrecinska (1932-2004) and Antoni Hrecinski (1929-2006), who, unfortunately, did not live long enough to see me graduate.
2 A Brief Introduction to Phrasal Verbs
To begin with, the constructions studied here are so called particle verbs, which are alternatively labelled as verb-particle constructions [Lipka (1972)], verb-adverb combinations [Kennedy (1920)], discontinuous verbs [Live (1965)], compound verbs [Palmer (1974/1976), Katamba (1993)], multi-word verbs according to Quirk et al. (1972; 1974, 1985) and Workman (1993/1996) as well as Claridge (2000) or Huddleston (2002/2005)]. They consist of a verbal base and a particle which may be an adverb or a preposition. Particle verbs constitute “a syntactic [and semantic] oddity in the language” to include the quotation by Darwin, Gray (1999:65). The diversity of labelling already conceives an idea how heterogeneous these verbs are. Their striking feature is that they behave like single items although they appear as two separate lexemes on the surface, as Leisi (71985:117), Fraser (1976), Quirk et al. (1972; 1985), Aarts (1989), or Wurmbrand (2000) observe.
Particle verbs are described in terms of three correlating aspects which flow in one another: the part of speech the particle belongs to, i.e. whether it is adverbial or prepositional, the syntactic behaviour of the entire combination and the meaning as well as the function of particle in particular context. The interdependence of these aspects makes the description of phrasal verbs difficult.
2.1 Phrasal Verbs: The Definition of Phrasal Verbs
The designation phrasal verb is generally used in two senses. Firstly, it serves as a global term for any kind of verb + adverb or preposition combinations , cf. Potter (1965) and Sinclair et al. (1989/2000), and secondly, in its narrow sense, to designate verb – spatial adverb combinations (cf. Kennedy  1967) such as put on, look up or take in and excludes prepositional verbs such as rely on, think about or depend on, as suggested by Quirk et al. (1972;1985), Palmer (1965) or Claridge (2000), inter alia.
Formally speaking, phrasal verbs consist, first of all, of monosyllabic and non-stative verbs of Germanic origin (cf. Potter 1965:287, Bolinger 1971:xi-xii, Brinton, Akimoto 1999:3), such as break, bring, call, cast, come, cut, do, fall, get, give, go, hang, hold, keep, kick, knock, lay, lie, live, look, make, move, pass, play, pull, push, put, run, send, set, sit, stand, stay, stick, take, talk, throw, turn (cf. Sinclair et al. 2000:vi) and adverbial particles aback, about, above, across, after, against, ahead, along, among, apart, around, as, aside, at, away, back, before, behind, below, beneath, between, beyond, by, down, for, forth, forward, from, in/into, of, off, on/onto, out, over, overboard, past, round, through, to, together, towards, under, up/upon, with, without (Sinclair et al. (2000:v) .
Their major characteristic is that they combine into units where the elements have given up their individual meanings to form a single lexical word (cf. Bolinger 1971:xii). Looking behind curtains, two problem comes into light:many adverbial particles, according to Heaton (1965:10) and O’Dowd (1998:18-19), have a corresponding preposition. This idea will be best illustrated on the basis of two entries from Cowie and Mackin’s phrasal verb dictionary. The verb look up has four different meanings and four different syntactic patterns (cf. Cowie and Mackin  1978:196-197).
As an intransitive combination with the particle as preposition, look up denotes either raise one’s eyes or improve, for instance:
(1) a. He didn’t look up from his newspaper when I entered the room.
b. Prospects for the small builder are looking up. (Cowie and Mackin  1978:196)
As a transitive combination with the particle in adverbial function, look up denotes either to search information about something or to search for somebody’s home so as to visit him:
(2) a. He comes back with an enormous dictionary, sits down and looks up the word.
b. I’ve promised that the next time I go to London I’ll look him up. (Cowie, Mackin 1978:196-197)
Claridge (2000:47) criticizes the common classification of phrasal verbs as too narrow:
Idiomaticity, after all, does not emerge out of nowhere, but is based in some way or other on the regular patterns of language. Literal phrasal verbs are the core from which figurative types are ultimately derived, and to which they are still connected by an identical, or in idiosyncratically frozen idioms at least similar, syntactic behaviour.
Further difficulty concerning this aspect is that the clear-cut boundary between literary and idiomatic expressions is not to be drawn easily. Therefore, together with Bolinger (1971), Cowie, Mackin (1978:x), Dixon (1982; 1992) and Palmer (1976), she (Claridge 2000:ibid.) claims that there is an idiomaticity gradient which divides phrasal verbs roughly into idiomatic, semi-idiomatic and literal particle verbs. This cline has following form:
I. Literal constructions where the meaning can be derived from both parts of the combination and their grammatical relation, e.g. John walked on the grass. (Dixon 1982:9).
II. Literal construction with a possibility to delete some parts of the prepositional phrase which can be understood from the context and socio-cultural knowledge, e.g. She put the rubbish out (of the building). (Dixon ibid.)
III. Constructions which are not literal anymore but involve a metaphorical extension in particular context but originate from a literal phrase, as it was shown above as you can run up a hill by moving upwards, it may be transferred onto increasing of the sum on the bill. A drowning man as well as a failing company may go under (cf. Dixon ibid.).
IV. Non-literal constructions which cannot be related to any literal combinations, e.g. They are going to have it out.(Dixon 1982:9-10).
Dixon (1982:11) states that in the course of time, particle verbs have followed a continuum as each stage flows into another, i. e. II flows into III and so on. It should be noted, in the spirit of this thesis that if the particle leaves its spatio-locative domain, it loses its full autonomy, to follow Kiffer (1965:3):“Homophonous preposition or adverb loses those qualities which make it either prepositional or adverbial and becomes simply a part of the pV [phrasal verb, B.H.]”.
2.2 Syntactic peculiarities
To refer back to the last section, the degree of semantic cohesion between the verb and particle correlates with the lexical status of particle which, in turn, affects the word order. This section goes outlines briefly some word-order alternations which are typical for adverbial particles to contrast them with full PPs. It is worth mentioning, that even among adverbial particles, the degree of semantic cohesion triggers some word order alternations, as observed by Dirven (1989b) and Cappelle (2004; 2005).
The pattern discussed here is the transitive verb-particle combination, V-Prt-NP which alternates with V-NP-Prt. This pattern is important as it reveals the unpredictability, even moodiness of particles in phrasal verbs. Moreover, this pattern helps to cull out phrasal verbs from verb plus PP combinations. The adverbial particle may occur on either side of the particle. Here it is not important whether the adverbial particle takes on a figurative meaning or whether it refers to location or motion, because both types of adverbial particles allow for word order alternations, cf. (3) – (6).
The sole condition is that the particle must move away from the verb if an unstressed pronoun appears. Erades (1961:57) justifies this by the role of discourse and stress:nominal objects are more strongly stressed than the pronominal objects and the latter are inclined to precede the particle whereas the former tend to follow them. Moreover, Erades (1961:57) states:
The principle governing the paces of the objects is (...) 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 (...) Erades (1961:57-58).
Pronouns as proforms, they stand for a noun that has been mentioned before or which is familiar to the hearer because it is immediately suggested by the situation. It is the absence of news value that determines the value that determines the place of the object (Erades 1961:58-59). In this respect, phrasal verbs and verb-adverb combination do not deviate (cf. Cappelle 2004:30-31).
(3) a. He gave in the document.
b. He gave the document in.
(4) a. He gave it in.
b. *He gave in it. (Radden 1989:49)
(5) a. Jill ran a big bill up.
b. Jill ran up a big bill.
(6) a. Jill ran it up.
b. * Jill ran up it (Cappelle 2004:31).
(7) a. Jill ran up a big hill.
b. *Jill ran a big hill up.
(8) a. *Jill ran it up.
b. Jill ran up it (Cappelle 2004:30).
Adverbial particles literal and non-literal enter word-formation which is not possible with prepositions, i. e. action nominalization (9)-(11). The verbs together with particle stay in adjacent position and are followed by a genitive – of.
(9) a . The taking out of the hostages (was not a clever move).
b. The pushing down of the piano (meant the end of their friendship ).
(10) a. *The taking to the front of the building of the hostages (was not a clever move).
b. * The pushing down the stairs of the piano (meant the end of their friendship) (Cappelle 2004:41).
(11) a. The looking up of the word was not easy.
b. The gluing up of the chair was not the best idea.
c. The prompt sending out of reports is recommendable (cf. Jackendoff 2002:72)
Modification by adverbs is, generally speaking, a criterion which determines the semantic cohesion between the verb and particle, cf. Bolinger (1971:11-12). The condition is that the phrasal verbs is not entirely idiomatic, (cf. Cowie/Mackin 1978:xlvi, Jackendoff 2002:71), and the adverb is not a manner adverb (cf. Darwin, Gray 1999:80). However, if the direct object intervenes between the verb and particle, the adverbial particle can also take specifiers, cf. (12). If the particle directly follows the verb, the adverbial must follow the nominal phrase otherwise it yields ill-formed results, as in (13) . Within the phrasal verbs, similar rule is true, (cf. Jackendoff 2002:ibid.).
Particle in post-object position
(12) a. I’ll look the answer right up.
b. Bill brought the wagon right back.
c. Fran put the model airplane right together.
Particle in pre-object position:
(13) a . I’ll look (*right) up the answer.
b. Bill brought (*right) back the plane.
c. Fran put (*right) together the airplane. (Jackendoff 2002:71)
If the bond between the verb and particle is too tight, following transformations are ruled out for phrasal verbs, the particle cannot undergo fronting (14), be coordinated (15), or repeated (16):
(14) a. Out jumped the frog! (Cappelle 2004:39)
b. *Up he made a story. (Darwin, Gray 1999:78)
(15) a. He carried the dinner in and out again
b . *He gave the paper in and out again. (Dirven 1989a:49-50)
(16) a. *I looked up your name, up her name, and up his name. (Darwin, Gray 1999:77)
b. I looked up one aisle, then up the next. (Darwin, Gray 1999:78).
Next group of transformation tests the constituency status of the particle and the following NP. Following transformations show that the particles do not form a constituent with the object but with the verb which is the case with adverbial questions, cleft sentences or preposition fronting, according to Cappelle (2004:30;38-39). This type of transformation is not available for semi-idiomatic and idiomatic particle verbs which show syntactic and semantic bond between both elements as it has been observed throughout this chapter. Some of the criteria are taken from Cappelle (2004:30-31):
(17) a. It was up a big hill that Jill ran.
b. *It was up a big bill that Jill ran. (Cappelle 2004:30-31)
Fronting of the PP:
(18) a. Up which hill did Jill run?
b. *Up which bill did Jill run up? (Cappelle 2004:30-31)
c. Up the syntactic tree moves the particle.(Cappelle 2004:38)
(19) a. This is the hill up which Jill ran.
b. *This is the bill up which Jill ran.( Cappelle (2004:30-31)
(20) a . Jill ran up a big hill and, Jack, up a small hill.
b :*Jill ran up a big bill and Jack, up a small bill. Cappelle (2004:30-31)
To draw temporary conclusion, if the particles and prepositions leave their literal or spatio-locative domain, they lose their lexical autonomy which is visible on word-order alternations such as pied-piping or preposition stranding in locative inversion or relative formation. These constructions are ruled out for particles in phrasal verbs.
3 Lexical status of particles in VPCs
3.1 Current Debate and Status quo
Last chapter has provided a brief insight into the heterogeneity of verb-particle constructions where the collocation exerts a considerable power on the combinations with the result that “although the postverbal string of words has the superficial appearance of a PP, it systematically fails all of the tests of constituency that PPs typically pass (Farrell 2005:96)”. It is true if the second element in question is a particle which doubles as adverb, e. g. about, above, across, after, along, around, by down, in, off, on, out, over, past, round, through, under, up, etc.(Quirk et al. 1985:1151)
It is worth mentioning that the particles, generally speaking, have a bad reputation among scholars, being termed as “wee wretched words” (Cappelle 2007) or being defined as “small words […] which do not easily fit into any clear word- class […]” or “an ‘escape (or cop-out) category’ for grammarians. ‘If it’s small and you don’t know what to call it, call it a particle’ […]” Hurford (1995:153)
In their literal meaning, particles are difficult to be assigned to a any lexical category, as they are hybrid between adverb and preposition, especially if they combine with intransitive verb of motion (Quirk et al. 1985:1151). Olsen (1998:316) asserts that:
Im Englischen dagegen scheint eine einzige Form in, out, off, down und dgl. als Präposition, Verbpartikel und Adverb fungieren zu können mit dem Ergebnis, daß es v.a. bei intransitiven Bewegungsverben schwierig ist, zwischen einer Partikel und einem Adverb zu unterscheiden.
The categorial problem between preposition and adverb is not new. Greenough et al. (1903:126) observe for Latin: “Particles cannot always be distinctly classified, for many adverbs are used also as prepositions and many as conjunctions” (Greenough et al. 1903:126)”.
The question which poses here is how to account for the particles in phrasal verbs which have undergone grammaticalization and assume aspectual function. From the perspective of traditional grammarians, particles in phrasal verbs are adverbial in nature as advocated by Kennedy (1920), von Schon (1977), and Cappelle (2004). The counter position is that particles should be treated, even only “by stipulation”, as intransitive prepositions as proposed by Emonds ( 1979), Jackendoff (1973), Dixon (1982), to name a very few because locative adverbs have not found their way in the framework of the generative grammar. This is largely confirmed by Lee (1998:134), who asserts, that the current language theory is based on a small number of discrete lexical categories which are defined by the clusters of morpho-syntactic properties which, however, require a very sharp and unambiguous assignment of lexical items to the relevant word class, which is not given here. The reason is, to follow Foskett (1991:34), the traditional grammar operates on the basis of functions which is of lesser importance within X’-bar syntax where endocentricity is leading principle. Correspondingly, does not take discourse and semantics into consideration which are significant for the description of particles in phrasal verbs.
In this framework, lexical categories are generally distinguished according to their distinctive features, i.e. whether they take verbal or nominal features [± N] and [±V] (Chomsky 1970). What this means is explained by König/Kortmann (1991:123):Categories with [+N] feature do not take bare NP objects, [-N] feature requires that the lexical categories are case assigners. Categories with [+V] can be modified by adverbs and those with [-V] form phrases that may be focussed on in cleft constructions.
Thereof, four major lexical categories are derived as follows:
(20) N [+ N, - V], V [- N, +V], A [+ N, + V], P [- N, - V]. It is important to fix that the lexical categories are determined by a set of distinctive features which determine their role in the syntagm. The absence of the features [± N] and [±V] is the driving force which leaves the adverbial particles without their proper category.
Emonds (1972/1979); Jackedoff (1973) suggest to enlarge the category of prepositions by including the adverbial particles to the class of prepositions and to treat pure adverbial particles like intransitive prepositions and the prepositional adverbs as those items which can be transitive and intransitive. The restriction is here that this analysis refers to spatio-locative terms which serve as adjuncts. Both authors, however, concede that their analysis cannot be mapped onto idiomatic particles. Huddleston (32005) applies this analysis onto the phrasal verbs, being well aware of the categorial and semantic restriction imposed on the phrasal verbs and nevertheless he attempts to save this line of argumentation. He starts from the suggestion that spatial adverbs or adverbial particles function as one-word phrases which are complement to the verbs (Huddleston 32005:280).
The opposed view is provided by Cappelle (2004, 2005) who as mentioned above, argues to grant the particles their own status because particles in phrasal verbs and prepositions vary considerably in their distribution, as pointed out in the section 2.2.
3.2 Towards A Solution to the Categorial Problem
3.2.1 Prepositions as Intransitive Prepositions
Generally speaking, prepositions they are heads which subcategorize for specifiers and complements. They yield following phrase structure rule (cf. Jackendoff 1973; Wollmann 1996, Rauh 1992; 1993):
(21) PP à(XP) P’ (YP), with XP as a specifier and YP as a complement.
The basic idea of Occam’s razor is introduced by Aarts (32008). This idea suggests that entities should not be multiplied unnecessarily which is interpreted as requiring that the simplest of competing theories be preferred to the more complex or that explanations of unknown phenomena be sought first in terms of known quantities (http://www.merriam-webster.com/dictionary/Occam%27s%20razor. Retrieved: March 3, 2009). That is the range of lexical categories should not be unnecessarily enlarged.
Emonds (1972/1979:187) starts from the fact that the noun phrases in VPCs are not objects of the particles. Moreover, directional adverbs and prepositions share the form and the inherent meaning (Emonds 1972/1979:188). The difference between those constituents is that in some cases, the occurrence of the direct object is optional. Similar opinion is shared by Jackendoff (1973), van Riemsdijk (1978), by Dixon (1982), Aarts (2001/2008) and Huddleston (2002/2005) who use this idea for their own analyses.
What distinguishes the preposition from the spatial adverb is that it takes obligatorily a nominal object. Members of group A and B may or may not take a nominal complement, confer:
(22) Chico ran into the opera house
in (Jackendoff 1973:345)
(23) The elevator operator kicked Groucho down the stairs
downstairs (Jackendoff 1973:346)
And there is anything to say against it as it is the discourse and semantic context which are responsible for the use of particular elements. The versatility of prepositional adverbs is triggered by their semantics. They carry the notion of “motion-through-location” and “terminus or result” (Bolinger 1971:86). Particle and the noun seem to share their labour as, according to Bolinger (1971:87), “a preposition embodies the motion feature and, the unmodified noun the terminus feature”. If the discourse requires clear delimitation of the path, the particle takes on prepositional function and if the limit is implied, the particle does not take complements in that it is semantically free and it can be analysed as preposition because this potential is latently available.
Fundamental justification is not to have to distinguish among various classes of particles according to their lexical category seems at first sight an attractive solution, especially with respect to the double function of prepositional adverbs (cf. Sroka 1972 or Quirk et al. 1985). According to Emonds (1979), claims that particles are generated by the base expansion rule which is can be realized in three different ways and this captures the latency of terminus within directional adverbs. He (Emonds  1979:189) suggests that like verbs, prepositions can be subcategorized for complements, thereof, he derives following subcategorization frames:
(24) with, +P, +__NP (at, for, toward, etc.)
in, +P, + __ (NP) (out, down, around, etc.)
apart, +P, +___ (away, back, together, etc.)
That is genuine prepositions are followed by a nominal complement whereas the purely adverbial particles do not subcategorize for nouns but may be followed by a full PP ( cf. Cappelle 2004:41), as it is the case with the phrasal-prepositional verbs, confer:
(25) a. break in on somebody’s conversation
b. catch up on my reading
c. catch up with somebody
d. check up on (‘investigate’)
e. come down with (a cold) (Quirk e al. 1972:817-818)
For semantic reasons, the class of directional particles is most versatile as it subcategorizes for optional nominal objects. Emonds (1979:188) suggests that the discontinuous structure is basic one and the continuous structure is derived via movement transformation:
(26) X + V-[P]PP -NPàX + V-NP-[P]PP (1979:189)
The reason is that only in this position; the former spatial adverbs can take be modified by intensifiers such as right see above, even if the combination is idiomatic (cf. 1972/1979:193).
There are certain problems with this approach. Dietrich (1960:165) observes with good reason that their use cannot be put into any kinds of general rules and the status of the particles depends largely on the verbs they combine with. Cappelle (2005:118-119) warns that:“[I] it is only when viewed in relation with the verb-particle combination as a whole that we can tell whether the particle is being used in a literal or in an idiomatic sense” (Cappelle 2005:118-119). In the context of phrasal verbs, one must be aware that the particles in phrasal verbs are not lexically autonomous prepositions according to in terms of Rauh’s (1993:121) criteria. Rauh (1992; 1993) distinguishes largely among two groups of prepositions, lexical prepositions, which determine these spatio-locative relations and non-lexical prepositions which fulfil largely grammatical functions or are selected by the verb, as it is the case with prepositional verbs. Lexical prepositions determine a two place argument structure and a theta grid with the external argument as THEME and internal arguments which are realized as LOCATION, SOURCE, PATH, GOAL to mark spatial relations (Rauh 1993:115, 119) which may or not be identified with the roles of verbs as it is in the case of adjuncts. They do not have features [-N,-V] which would render them fully prepositional. Therefore, they do not assign Case to the nominal complement because they do not take complements; they do not have proforms such as here or there. Moreover, they do not undergo movements which are exposed in the section 2.2., to name a few of the criteria. Moreover, particles can enter word-formations which is ruled out for prepositions.
Second evidence is that in their inherent semantics, particles in phrasal verbs refer to motion through location which can be decomposed as motion through location with terminus and result, even in the abstract domain. However, there are many resultative phrasal verbs which do not refer to any kind of motion anymore. Next section points to the counter-position as proposed by Cappelle (2005).
 In his works, Cappelle (2004, 2005), has compiled the entire range of features which have been found in Legum (1968), Bolinger (1971), Lipka (1972), Quirk et al. (1972), Fraser (1974/1976), Lindner (1983), Kayne (1985), Radford (1988:90-101), O’Dowd (1998), Haegeman/Guéron (1999:253-260), Claridge (2000) and Huddleston (2002:271-290), Dehé (2000; 2002.) and Elenbaas (2006), to name only a few.
 If the preposition is lexically autonomous, a.it takes complements and specifiers;
b. it has a semantic proforms for P;
c. it undergoes category specific movements and can be coordinated;
d. it assigns Case;
e. it determines semantic projection properties, i.e. P is involved in semantic well-formedness conditions within Pmax;
f. it determines the syntactic proforms from semantic point of view;
g. it has an argument structure and a theta grid;
h. it defines the properties of the arguments, of the theta roles and of the relation between the arguments (Rauh 1991: 121)