Trial and Error? - About Past and Current Approaches Towards Idiomaticity

Table of Contents

1 Introduction

2 Makkai´s and Strässler´s Views on two earlier approaches towards Idiomaticity
2.1 Katz and Postal
2.2 Chafe

3 Gibb´s Views on the Analyzability of Idioms

4 Conclusion

Appendix

Bibliography

Introduction

"People are not considered competent speakers of a language until they master the various clichéd, idiomatic expressions that are ubiquitous in everyday discourse." (Munro, 1989 as cited in Gibbs, 1995: 97)

Idioms play a significant role in the natural use of not only the English but every language.

Whenever one wants to achieve the level of a native speaker it is absolutely indispensable for him or her to internalize the use of these expressions. Apparently, the meanings of the words used in an idiom often do not deliver enough knowledge to decipher the correct meaning of the whole expression. This is a major point that makes the acquisition so difficult for foreign learners.

This paper will, of course, not be used to deal with the topic of idioms as a whole. Idiomaticity represents too large a field, in which entire studies can deal with one single idiom. Instead, my goal is to deliver an idea of what idioms are and ´scratch the surface`, so- to-speak, of the incredibly wide and always changing field of idiomaticity. A major point that is to be discussed, deals with the actual analyzability of idioms. As mentioned before there is often a kind of incongruity between the expression and its single parts. In this part the focus will be on Raymond Gibbs´ studies concerning this subject. He considers many aspects involved in the topic of analyzability and refers to the viewpoints of many other linguists, hence a lot of different ideas will be discussed and also criticized. Perhaps, there is even more of an explanation given in some cases by the constituents of the idiom than one usually expects.

While Gibbs´ dealing with analyzability is one of the more current studies on idiomatic aspects, another part of this paper will rather give an insight on past approaches which were published in succession in the 1960´s. Because idiomaticity is a sector of linguistics that is constantly changing and developing new prospects will always be discovered, but nevertheless preceding studies, for example by the here presented Katz, Postal, and Chafe shall not be forgotten.

However, first of all it has to be agreed on a definition for the term ´idiom`. As one might expect there have already been innumerable approaches in various different manners towards this little word. Hockett in 1956, for instance, pointed out: "There is no upper limit to the size of an idiom; one can validly conceive of a whole poem, or novel, or the King James version of the Bible, as a long idiom" (Hockett 1956: 222). By this statement one is able to see how far a discussion even about the definition may go. One appropriate definition, in my opinion, is the one given in NTC ´ s American Idioms Dictionary (1994: iv):

"All languages have phrases or sentences that cannot be understood literally. Even if you know all the words in a phrase and understand all the grammar of the phrase completely, the meaning may still be confusing. A phrase or sentence of this type is said to be idiomatic. Many proverbs, informal phrases, and common sayings offer this problem."

The idea of literal meaning, as referred to in this definition, will also be one of the significant points discussed throughout the paper. As a conclusion for this study the presented approaches will be compared to each other and this will probably give a little hint about the development of studies concerning idiomaticity over the last decades.

2. Makkai´s and Strässler´s Views on two earlier approaches towards Idiomaticity

2.1. Katz and Postal

In 1963 the linguists Katz and Postal used a short article called ´Semantic Interpretation of idioms and sentences containing them` for "the first attempt to deal with idioms within a transformational grammar", as Strässler remarks (1982: 30). To begin with, it has to be mentioned that the two linguists made a general distinction between two types of idioms, lexical idioms and phrase idioms. In general, idioms are defined as follows (Makkai 1972: 47):

"The essential feature of an idiom is that its full meaning, and more generally the meaning of any sentence containing an idiomatic stretch, is not a compositional function of the meanings of the idiom´s elementary grammatical parts."

At this point Katz and Postal are criticized by Makkai for leaving out "such obvious (lexical) idioms as hot dog (or) man-of-war,..." (Makkai 1972: 47). He recognizes these idioms as being "genuine, multi-word lexical idioms" (Makkai 1972: 47), instead the idea of lexical idioms is illustrated with the words baritone and telephone, here consisting of the single parts bari + tone and tele + phone. These two as he also points out have a "debatable idiomatic status" (Makkai 1972: 47). The most disapproved aspect, however, is that the different facets of lexical idioms are not dealt with, as Makkai for example says: "No attempt is made at classifying idioms" (Makkai 1972: 47). Strässler agrees in that "lexical idioms are of less interest for Katz and Postal" (Strässler 1982: 30). On the contrary, "their paper is mainly concerned with phrase idioms" (Strässler 1982: 30). As Katz and Postal focused on phrase idioms one would expect both Strässler and Makkai to recall some of the examples Katz and Postal need to have used for the explanation of their concept of phrase idioms. The distinction between lexical and phrase idioms hereby remains fairly obscure. However some credit has to be given, at least to Strässler, as he provides the reader with what is supposed to be a definition of lexical and phrase idioms. In my opinion, Katz and Postal´s definition seems very difficult to follow, especially without emphasizing examples. Strässler quotes their definition (Strässler 1982: 30):

"The characterization of an idiom as any concatenation of two or more morphemes whose compound meaning is not compositionally derived from the meanings of the concatenated morphemes does not differentiate those idioms that are syntactically dominated by one of the lowest syntactic categories, i.e. noun, verb, adjective, etc., from those whose syntactic structure is such that no single level syntactic category dominates them. Let us call the former type ´lexical idioms` and the latter ´phrase idioms`." (Katz, Postal 1963: 275-6)

However, the results of Katz and Postal´s studies conclude in their suggestion to divide "the dictionary into two parts, a lexical-item part and a phrase-idiom part" (Strässler 1982: 30). The advantage of this division is explained as follows: "This method enables them to assign readings to higher level constituents in an underlying phrase marker, and not to terminal symbols" (Strässler 1982: 30).

Furthermore, Katz and Postal discuss the transformational deficiencies of idioms. By presenting the active and passive form of an idiom (to kick the bucket) they distinguish between a phrase-form that is understood ´idiomatically` (see <1>), and a phrase-form that only carries a literal meaning (see <2>) (Strässler 1982: 31).

<1> John kicked the bucket.

<2> *The bucket was kicked by John.

One aspect that could be criticized at this point is that Katz and Postal only dealt with one idiom; hence, one cannot assume that the difference between idiomatic and literal understanding can be generally used for any idiomatic phrase. But, as a conclusion Strässler remarks that "we should bear in mind that they (Katz and Postal) were first to suggest a separation of the dictionary into a lexical-item part ... (and) ... to point out transformational deficiencies as well as a possible way to overcome them" (Strässler 1982: 31).

2.2 Wallace Chafe

In 1968 Wallace Chafe´s article entitled ´Idiomaticity as an anomaly in the Chomskyan paradigm` was published. As Strässler comments, "he takes idiomaticity as an example of how one might criticize Chomsky´s view of language and demonstrate his concept of generative semantics" (Strässler 1982: 34). Both, Makkai and Strässler, agree that Chafe follows Thomas Kuhn´s argumentation from Kuhn´s well-known book ´The structure of scientific revolutions` (1962). Makkai recalls this central idea which "convincingly demonstrates that transformational-generative grammar cannot handle meaning in general, and idiomaticity in particular" (Makkai 1972: 51). Coming back to the criticism Strässler points out "four different features of idioms that have to be considered as anomalies in the Chomskyan paradigm". These are as follows (Strässler 1982: 34):

1. The meaning of an idiom is not an amalgamation of the meanings of its parts.
2. Most if not all idioms exhibit certain transformational deficiencies.
3. Some idioms are syntactically ill-formed.
4. Any well-formed idiom has a literal counterpart, but the text frequency of the idiom is usually much higher.

As Makkai further exhibits Chafe reveals "the trouble (which) lies with the transformationalists´ insistence that syntax is generative and semantics and phonology merely interpretive" (Makkai 1972: 51). Chafe also refers back to Katz and Postal´s analysis of passivization (see above) which sometimes turns out to lead to a "loss of idiomaticity in these forms, (and) sometimes not". However, he also states that both interpretations could as well sometimes be thinkable (Makkai 1972: 51). Coming back to the "four peculiarities of idioms"

(Makkai 1972: 51), according to Chafe each of them needs to be "explained by a theory of language adequate to cope with idiomaticity" (Chafe 1968: 111-112). One strong contrast between Chafe´s and Chomsky´s approaches is that "Chafe puts semantic structures into the base component at the beginning of a derivational process" (Strässler 1982: 35). Chafe is of the opinion "that semantic structures are not immediately presented, since two processes have to operate first". For these processes the terms ´linearization` and ´symbolization` are used. First of all Chafe proceeds on the assumption that one deals with three types of structure: the semantic structure, the surface structure, and the phonetic structure. To elucidate this, we will follow an example which is also presented in Strässler´ s study (1982: 36). Assumed is that the semantic structure is cat/plural, the surface structure is cat + plural, and the phonetic structure as a result is cats. The process of ´linearization` takes place between the semantic and the surface structure, while the process of ´symbolization` occurs between the surface and the phonetic structure. At first glance this concept by Chafe may not appear as being very distinct to the reader, but Strässler adds good support for Chafe´s concept. He points out that one encounters two possible changes: a "semantic change" due to the concept of linearization and a "phonological change" on the basis of the concept of symbolization (Strässler 1982: 36). It is further made clear that these concepts have to be slightly adjusted for a use on idioms as "idiomaticity creates new concepts with symbolizations which already exist" (Strässler 1982: 36). For this instance "the semantic unit is postsemantically changed into another unit" (Strässler 1982: 36). After that, the symbolization process is just the same as the one presented in the first example (cats). However, there is of course also an extra example given for easier understanding. If one considers the example white elephant here white does not carry the same meaning as does the color white. In fact the meaning is special when the word occurs in this particular phrase, next to elephant. So instead of the process of ´linearization`, in this case one has to deal with a sort of ´postsemantic` process. This idea helps to show differences between several meanings of the same word, often depending on the compound of a phrase or sentence. Unfortunately Chafe does not actually explain the phrase `white elephant´. So one might see it as an ordinary two-word phrase, especially because Chafe does not mention that it is de facto an idiomatic expression. According to NTC´s American Idioms Dictionary (1994: 390) `white elephant´ describes "something that is useless and which is either a nuisance or is expensive to keep up".

However, Chafe takes his model of semantic processes another step further when he explains "that a configuration can be understood idiomatically and literally at the same time" (Strässler 1982: 37). Once again ´to kick the bucket` is used to illustrate the argument. It carries not only

the idiomatic meaning of `to die´ but also the literal meaning `to kick the bucket` (the same as to hit the pail), though one rarely uses it to say that someone physically kicks a bucket (Strässler 1982: 37). A reason may be the near-exclusive use of the phrase in the idiomatic sense, and hence the inherent connotation of `to die´.

Moreover, Chafe´s model offers an "explanation for transformational deficiencies"(Strässler 1982: 37). Chafe depicts "that ´kick the bucket` (...) is subject to the same inflectional modifications as ´die`" (Chafe 1968: 122). This is underlined by the possible modifications "Sam has kicked the bucket" and "Sam may kick the bucket" (Chafe 1968: 122). But Chafe urges that "we cannot (...) inflect parts of the literalization of this idiom, for those parts are not present at the semantic stage where sentences are generated" (Chafe 1968: 122). This means that one cannot change, for example the form of bucket into bucket s, because it "is introduced postsemantically by a literalization rule which specifies that it must be singular and definite" (Chafe 1968: 122). In the same way one cannot add an adjective to a phrase, e.g. "Sam kicked the wooden bucket" (Chafe 1968: 122). Even though this sentence is not wrong, the idiomatic sense of ´kick the bucket` is absolutely taken away and thereby only the literal meaning prevails.

The aspect of passivization, previously discussed by Katz and Postal, is also considered by Chafe. For this, the linguist even defines a new term. If passivization is possible, as with the idiom ´bury the hatchet`, he would "perhaps (...) distinguish a class of ´semi-idioms`" (Chafe 1968: 122). These semi-idioms´ literalizations would then be "amenable to restricted kinds of semantic tampering, including passivization" (Chafe 1968: 122).

At this point his attempt to standardize this model of semi-idioms has to be criticized. Once again the problem occurs that linguists tend to generalize concepts after considering some idioms (or even only one). When a point of conflict occurs they often modify their whole idea to avoid any problems. Ironically, Chafe himself criticizes just that by quoting Kuhn (in Strässler 1982: 38): "Instead they devise numerous articulations and ad hoc modifications of their theory in order to eliminate any apparent conflict". As Strässler also points out, with the idea of semi-idioms, "he (...) falls into exactly the same mistake" (Strässler 1982: 38).

After considering Katz´ and Postal´s and Chafe´s approaches towards idioms one might conclude that every linguist back then tried to classify idiomatic expressions in some way. As it was presented these attempts were not very successful, because the firmness of the different lines of argumentation seemed to vanish when an examined idiom deviated from the rule that had just been made. One question that comes up is whether idioms are classified now, more than thirty years after these approaches. This question cannot be answered satisfactorily. Idiomaticity is and may always be a part of linguistics that tends to show the opposed opinions of different linguists and leaves enough room for numerous viewpoints as well as the inevitable discussions.

3. Gibb´s Views on the Analyzability of Idioms

The question that commonly comes up when discussing the analyzability of idiomatic expressions is whether it is preferential to deduce the meaning of the expression as a whole or from the individual meanings of its constituents.

Raymond Gibbs distinguishes between "the almost completely analyzable idioms" and "the much less analyzable idioms" in answering that very question (Gibbs 1995: 98). If none of the single words in an expression shows a connection to the final meaning of the idiom, he then uses the phrase "arbitrary stipulation" (Gibbs 1995: 99). This "unmotivated interpretation" (Gibbs 1995: 99), as it is also called, is one of the many problems that arises when one takes a traditional view of idioms. The often used example of ´ kick the bucket ` shall explain this idea as kick, the, and bucket all do not necessarily pertain to "the act of dying" upon first glance (Gibbs 1995: 99). According to Gibbs, this famous phrase could have "an obscure historical reason" (Gibbs 1995: 99), but native speakers normally are not aware of the origins of idioms when they use them, and in fact they may even "be completely ignorant" (Gibbs 1995: 99). That people often use the idioms without caring about their origin or even about the single words of it may be further demonstrated by the following example. In the United States of America people use the idiom ´ to toe the line ` in order to say that someone is close to the edge or is in danger of going too far. When asked to write the idiom, many of the people have difficulties and incorrectly write *`to tow the line´¹. Even though it is an ubiquitous phrase and used frequently in everyday language, there are persons who do not even know the meaning of its constituents.

But there are other sorts of idioms. These can be analyzed to a certain extent (at least) by analyzing the parts of the expression. One example Gibbs offers in this context is the phrase ´ spill the beans `. Speakers, as he points out, do know "that beans refers to an idea or secret whereas spilling refers to the idea of revealing a secret" (Gibbs 1995: 100). Even simpler is the example of ´ blow your stack ` as the word blow can be put into a direct relation to "the act of suddenly releasing or expressing internal pressure from the stack or from the human body or mind" (Gibbs 1995: 100). This second group, in other words, is easier understood because the single words of the phrase already have established a sort of connotation in the speaker´s or learner´s mind. But as mentioned before, the extent to which idioms can be analyzed can largely differ as there is a great scope between ´ kick the bucket ` and ´ blow your stack `. Another belief widely held among many linguists and psychologists is that readers or listeners switch to a different mode of comprehension when they encounter idiomatic expressions. More specifically, they switch from "a normal, literal mode of processing to a more specialized non-literal mode of processing" (Gibbs 1995: 100). Gibbs strongly rejects this belief. He indicates that one of the most common mistakes made is when one assumes that distinguishing between literal and non-literal (also called figurative) meanings of idioms is an easy task. To underline his statement he presents different examples that show the vagueness of the contrast which seems to be taken for granted (Gibbs 1995: 101):

"For example, the contrast between idioms and their literal meanings, metaphors and their literal meanings, metonymies and their literal meanings, ironic statements and their literal meanings, provides very different notions of literal meaning."

Gibbs, however, does not go beyond that statement and does not provide any really convincing example. Although he asserts that ironic statements and metaphors provide different notions about their prevailing literal meanings, this feature is not proven at any point. As a result, his argumentation concerning this aspect, to use an idiom myself, ´ is standing on shaky legs`.

However, at this point, Gibbs is trying to alert us to one major problem which is the interpretation of the concept of `literal meaning´. Literal meaning, as Gibbs quotes, can mean "conventional meaning, context-free meaning, truth-conditional meaning, subject-matter meaning" (Gibbs et al.: 1993; Lakoff: 1986). All these categories of meaning that Gibbs mentions are unfortunately not supported by examples from his or Lakoff´s previous study. As Gibbs further remarks, people apparently apply "different senses" in "different ways" (Gibbs 1995: 101). These senses depend on the kind of utterance, the context, and the task (Gibbs 1995: 101). In one of his earlier studies, Gibbs also refers to psychological studies which demonstrates that "people do not access the same invariant, literal meanings each and every time they encounter a word in spoken or written discourse" (Gibbs 1995: 101). As a conclusion, Gibbs states that one cannot assume that parts of idioms or even entire idiom phrases have "easily determined literal meanings" (Gibbs 1995: 101). He intensifies this by expressing that "people do not have an adequate sense" (Gibbs 1995: 101) for literal or word meaning. These last two terms are even emphasized by him as being different notions.

Many scholars may find this viewpoint concerning literality and idiomaticity slightly disturbing. Here the problem is, as Gibbs identifies, that scholars "continue to adhere to the idea that word meanings can be defined in some context-free manner" (Gibbs 1995: 101). Gibbs again rejects this seemingly ubiquitous belief. His "suspicion" would rather be (similar to Nunberg: 1978) "that idioms differ quite a bit in terms of how their parts contribute to their meaning as a whole" (Gibbs 1995: 101). Gibbs next clarifies this facet with the example of ´ pop the question `. While the word ´question` does have a general meaning (the common use in everyday discourse), there is also a slightly changed meaning to it in the idiomatic use. Here it is rather restricted, so-to-speak, as its only meaning is to propose to someone for their hand in marriage. Gibbs then refers to various study results (Cacciari: 1993; Flores d` Arcais: 1993) concerning the problem in which way "the meanings of individual words (e.g. literal, conventional, metaphorical) are accessed during the processing of familiar and unfamiliar idioms" (Gibbs 1995: 101). In some of these studies it is revealed that many idioms actually "have key points or uniqueness points", or in other words, "places at which they become uniquely identifiable" (Cacciari, Tabossi: 1988; Tabossi, Zardon: 1993). Unfortunately, Gibbs does not offer the reader any excerpts or perhaps statistics from the studies mentioned which would have helped in following his line of argumentation. Instead he confined himself to focusing on their general ideas. Even though these different study results, according to Gibbs, provide some evidence "that is consistent with the data on the analyzing of idioms", he urges that in both, linguistics and psychology, there is still a great deal of knowledge to be discovered. More should be found out about the "different types of word meanings that play a role" not only "in the linguistic behavior" but also "in the learning, use, and understanding of idioms" (Gibbs 1995: 102).

Nevertheless, there are a number of interesting linguistic and behavioral consequences to be presented. These consequences, in fact, may reflect the previous idea on how idioms differ in their degree of analyzability. In one particular series of studies, we become aware "that the semantic analyzability of an idiom affects people´s intuition about its syntactic productivity" (Gibbs, Nayak: 1989). One example of this would be that "people find semantically analyzable or decomposable idioms more syntactically flexible than unanalyzable idioms" (Gibbs 1995: 102). This aspect is made clear by the example of ´John laid down the law`. When the structure of this sentence is changed into the passive form ´The law was laid down by John`, the figurative meaning is not disrupted and hereby the original meaning remains. However, "semantically unanalyzable idioms tend to be much more syntactically frozen"

(Gibbs 1995: 102). This idea of so-called `frozenness´ refers back to the well-known concept of Bruce Fraser´s "frozen hierarchy" (1970). In his theory Fraser subdivides the transformability of idioms into seven levels. Each of those levels describes a certain `frozenness´ (meaning rigidity of form) of an expression. The range of levels ranges from complete frozenness (Level 0) to totally unrestricted expressions (Level 6) (Strässler: 1982: 38-39)². As an example of frozenness Gibbs chose ´John kicked the bucket` versus ´The bucket was kicked by John`. The passive form in this case, though not syntactically false, has a disrupted figurative meaning and would consequently never be used in this way. Regrettably, Gibbs does not refer to Fraser´s original concept in more detail. He rather uses the term ´frozenness` as if it was self-evident which is in my opinion not very helpful and the readers might even have some problems with the actual idea of `frozenness´. At least non- native speakers will find it problematic to assign phrases to the different levels of the hierarchy.

As he moves on in his argumentation Gibbs points out that analyzability also plays a significant role for the "immediate online interpretation" (Gibbs 1995: 102). Therefore, a closer look at the contribution to the individual components of an idiomatic phrase is essential. Hence, we will once again consider the two previously used examples. In ´lay down the law`, which is, as we already know, an analyzable idiom, the individual components "systematically contribute to the figurative meaning" of the phrase, as Gibbs quoted from Peterson & Burgess(1993) (Gibbs 1995: 102) and he continues furthermore with "people process idioms in a compositional manner in which the meanings of the components are accessed and combined according to the syntactic rules of the language" (Peterson & Burgess: 1993) (Gibbs 1995: 102). In contrast to that, if one tries to analyze a semantically unanalyzable idiom (e.g. ´kick the bucket`) in a strictly compositional way, the result will not provide much information about the prevailing figurative meaning.

Undoubtedly, there are some major difficulties concerning an understanding of unanalyzable idioms. Thus, two questions arise: Why do people understand these seemingly complicated and unanalyzable idioms? and How do they understand them?

According to Gibbs (1995: 103):

"Understanding unanalyzable idioms requires that people first do some sort of analysis where the individual words are examined to see if they have independent meanings that contribute to the meaningful interpretation of the idiom as a whole."

If "this process fails to produce an acceptable interpretation in a context", Gibbs suggests that people will probably recapture "the conventional, figurative meanings of these phrases from their own lexicons" (Gibbs 1995: 103). He bases this conclusion about the comprehension of idioms on so-called ´reading-time studies` (Gibbs, Nayak, & Cutting: 1989) which according to him brought to light that people "took significantly less time to process (...) analyzable idioms than to read unanalyzable expressions" (Gibbs 1995: 103).

As Gibbs had antecedently found out in three other studies (1980; 1985; 1986), people do not "automatically" cipher the "literal, context-free interpretations of idioms " (Gibbs 1995: 103). Another result that was uncovered stresses that there is no recognizable difference among idioms based on the form in which the idiom is built. For instance, an ill-formed idiom like ´pop the question` is just as easily understood as a well-formed phrase (e.g. ´kick the bucket`). But how then did these difficult idioms gain equal status as analyzable idioms in terms of comprehension?

To learn the answer, the processing behavior in a person's childhood should be investigated because the language including all their specificities (like idioms) develops very strongly in the early years of one´s life. Gibbs emphasizes that "children actually experience greater difficulty learning the meaning of semantically unanalyzable idioms" (Gibbs 1995: 103). And the reason for that brings us right back to our very first and most common assumption. These difficulties, according to different studies (Gibbs:1987, 1989; Nippold and Martin: 1989), present themselves "because these phrases´ non-literal interpretations cannot be determined through analyses of their individual parts".

This then suggests that the individual parts do play an important role in idiom comprehension. To sum up this part about analyzability, it can generally be said that a certain hint (or even more) concerning the overall meaning of an idiom may be given by the constituents of the phrase. However, we cannot dismiss numerous other idioms that do not seem very reasonable. These phrases, in most cases, have developed from an unknown historical heritage. In comparison, they are more difficult to learn and understand than other idioms, but nevertheless they are frequently used in everyday language. My assessment at this point would be that the acquisition of these abstract idiomatic phrases simply goes hand in hand with habitual use in a native-speaking environment. Only there can one attain knowledge concerning the correct meaning and the correct use of any idiomatic expression.

4. Conclusion

After considering three different approaches towards idiomaticity these should now be put into perspective. One obvious difference between the early approaches (which means Katz´/Postal´s and Chafe´s) and the recent ones (e.g. Gibbs) can be indicated. This is that Gibbs´ study offers a greater number of idioms that is exercised which always helps in underlining the argumentation. Contrary to that, the other two approaches that were presented here tended to generalize after considering only one or two examples. But of course, this does not mean that all the early approaches have failed to create a certain development in revealing more and more aspects in the field of idiomaticity.

But there is also a great similarity which catches the reader's eye. The aspect of passivization seems to be as much a topic as it was thirty years ago. In Katz´ and Postal´s studies the active and passive forms of an idiom are presented in order to show the distinction between idiomatic and literal meaning (also see page 5). This is as well suitable to show the problem of considering only one example, as it was mentioned before. In the other study Chafe also discusses passivization and is rather vulnerable to attack because of the lack of examples. In fact his argumentation is in danger to lose validity when Chafe comes across a syntactically different example (also see page 8).

Gibbs, later, compares an active and a passive form of idioms to find out that sometimes the figurative meaning of the expression may be disrupted. From there he turns to the idea of ´syntactical frozenness` by Arnold Fraser. This approach by Fraser is also about thirty years old and shows how different theories are based on one another. Critics, though, could converse this and remark that there has not been much development as linguists are still talking about passivization, as they already did many years ago. The question of development in idiomaticity, it seems, remains more or less a matter of opinion.

Bibliography

Cacciari, C. (1993). The Place of Idioms in a Literal and Metaphorical World. IN Cacciari, C. and P. Tabossi (eds.). Idioms: Processing, Structure, and Interpretation.

Cacciari, C. and P. Tabossi(1988). The Comprehension of Idioms. Journal of Memory and Language 27, 668 - 683.

Chafe, W. (1968). Idiomaticity as an Anomaly in the Chomskyan Paradigm. Foundations of Language 4, 109 -127.

Flores d´Arcais, G. (1993). The Comprehension and Semantic Interpretation of Idioms. IN Cacciari, C. and P. Tabossi(eds.). Idioms: Processing, Structure, and Interpretation.

Fraser, B. (1970). Idioms within a Transformational Grammar. Foundations of Language 6, 22 - 42.

Gibbs, R. (1980). Spilling the Beans on Understanding and Memory for Idioms in Conversation. Memory and Cognition 8, 449 - 456.

Gibbs, R. (1985). On the Process of Understanding Idioms. Journal of psycholinguistic Research , 465 - 472.

Gibbs, R. (1986). Skating on Thin Ice: Literal Meaning and Understanding Idioms in Conversation. Discourse Processes 9, 17 - 30.

Gibbs, R. (1987). Linguistic Factors in Children ´ s Understanding of Idioms. Journal of Child Language 14, 569 - 586.

Gibbs, R. (1989). Understanding and Literal Meaning. Cognitive Science 13, 243 - 251.

Gibbs, R., Nayak N., and C. Cutting (1989). How To Kick the Bucket and Not Decompose: Analyzability and Idiom Processing. Journal of Memory and Language 28, 576 - 593.

Gibbs, R. et al (1993). Literal Meaning and Figurative Language. Discourse Processes 16, 387 - 404.

Gibbs, R. (1994). The Poetics of Mind: Figurative Thought, Language, and Understanding. Cambridge University Press, New York.

Gibbs, R. (1995). Idiomaticity and Human Cognition. IN Everaert, Martin et al. (eds.).

Idioms: STRUCTURAL and PSYCHOLOGICAL PERSPECTIVES.

Hockett, C.F. (1956). Idiom Formation. IN For Roman Jakobson. Essays on the Occasion of his Sixtieth Birthday, 222 - 229.

Katz, J.J. and P.M. Postal (1963). Semantic Interpretation of Idioms and Sentences

Containing Them. M.I.T. Research Laboratory of Electronics; Quarterly Progress Report 70, 275 - 282.

Lakoff, G. (1986). The Meanings of Literal. Metaphor and Symbol Activity 1, 291 - 296. Makkai, A. (1972). Idiom Structure in English. The Hague: Mouton.

Nippold, M. and S. Martin (1989). Idiom Interpretation in Isolation Versus Context: A

Developmental Study with Adolescents. Journal of Speech and Hearing Research 32, 59 - 66.

Nunberg, G. (1978). The Pragmatics of Reference. Indiana University Linguistics Club, Bloomington.

Strässler, J. (1982). Idioms in English: A Pragmatic Analysis. Tübingen: Gunter Narr Verlag.

Spears, R. A. (1994). NTC´s AMERICANS IDIOMS Dictionary. Chicago: NTC Publishing Group.

Tabossi, P. and F. Zardon (1993). The Activation of Idiomatic meaning in Spoken Language Comprehension. IN Cacciari, C. and P. Tabossi (eds.). Idioms: Processing, Structure, and Interpretation.

[...]

¹ Information retrieved through personal eMail correspondence with a native speaker (also see Appendix)

² It should further be said that Fraser´s transformational hierarchy is not universal. As Burger found out in 1973 it does not apply for every language, e.g. German. (Strässler 1982: 40).