English Relative Clauses in Dialects of the GB Isles - Essay Example

In the journal article “English Relative Clause Constructions,” by Ivan Sag, a novel theory of clause construction is posited. This theory is a complex amalgamation of different constructs of the relative clause. In the article, “Relative Clauses in Dialects of English,” Tonja Herrmann examined the dialects of six distinct regions of the British Isles, and applied long-standing hierarchical theory to these regional dialects. This paper will attempt to compare and contrast the two theories, along with synthesizing them. In part one, the basic premises upon which Ms. Herrmann builds her theory will be examined. In part two, I will examine the basic premise upon which Mr. Sag builds his theories. Part three will examine Ms. Herrmanns theory of grammatical construction and part four will examine the commonalities that the two theories have, by paying special attention to the concept of pied-piping. To begin, one must examine the basic premise upon which Ms. Herrmann built her grammatical constructs. In the “Relative Clauses in Dialects of English,” the author describes an extensive study of the British Isles. She divided the Isles into six broad sections - Central Midland, Central north, Central Southwest, East Anglia, Northern Ireland and Scotland. She concentrated on the prototypical relative clauses, or adnominal clauses, and she compared these clauses across the different dialects of the sections listed above, with an eye towards identifying the features that have become commonplace across different dialects, and also identified how the commonalities and differences across different dialects identified certain dialects as more closely related to Standard English than other dialects. Her study also can identify where different dialects intersect, which is known as dialectical levelling. (Herrmann 22). In the process, she also identified broad characteristics of the various dialects she studied. Central southwest, Central North and Northern Ireland have broad dialectical speech, ie, the speakers in these regions tend to speak in one dialect, while the speakers in East Anglia, Central Midland and Scotland have more heterogenous speech. (Herrmann 24). Central Southwest and Northern Ireland are almost entirely broad speakers. Broad speakers tend towards more non-standard features, which is gradually transforming into traditional features, and these are influencing Standard English as a whole. (Herrmann 22). The basic findings of the study were that the relative particles (zero, that, what as) outnumbered relative pronouns (who, which, whom, whose) 3:1. (Herrmann 24).  “That” is most popular (39%), followed by “zero” (28.1%), “which” (15.1%), “who” (10.1%), nonstandard “what” (6.8%), and “as” (.8%). “Whom” (.2%) and “whose” (.1%) are “very unusual in dialectical speech.” (Herrmann 24). The popularity of “that” increases the dialects go northward. In the Central Southwest, and East Anglia, the marker “that” occurs roughly 25% of the time. In Central Midland, that percentage jumps to 40.3, 43.5 in Central North, 46.5 in Scotland, and 50.1% in Northern Ireland. (Herrmann 27). The non-standard markers, “what” and “as” are only popular in the south, with “what” being the substantially more popular of the two. “What” is found in 22.3% in the Central Southwest, and 15.9% in East Anglia. Once you get north, the percentages drop to around 6% in the Central Midlands, 2.5% in the Central North, .4% in Scotland, and .5 in Northern Ireland. But “as” isn’t found in the south at all, and is weak elsewhere - 2.4% in Central Midlands, 1.4% in Central North, and not found in Scotland or Northern Ireland. (Herrmann 28). The relative presence of the number of -wh words denote a region that is closer to Standard English than other regions, because -wh markers are a hallmark of Standard English. The study bears this out, as East Anglia, which is closest to standard of the regions studied, has a substantial number of -wh pronoun instances 41.8%, compared to 2.4% in Northern Ireland. Central north has 18.7% of -wh pronouns, central southwest 22.3% of -wh pronouns, Scotland 29.9 and Central Midlands 33.8.(Herrmann 28). Another finding was that the regions who used more non-restrictive clauses than other regions were the regions whose language more closely resembled Standard English. (Herrmann 38). What she found was that restrictive clauses were four times as frequent as non-restrictive clauses (80.1% to 19.9%). (Herrmann 38). “Which” is the majority choice for nonpersonal non-restrictive clauses with 61.9%, and personal non-restrictive clauses primarily used “who” (14.8%). Pages 38-39. “What” is used in 11.4% of all non-restrictive clauses. Page 38. “Zero” makes up 2.7% and that appears 3.8% of the time. (Herrmann 38). “What” is not gender specific, so can be used for either personal or non-personal clauses, yet used 84.2% of time in a non-personal clause. “Which” is used in the same manner as what in non-personal clauses, and the study shows that the languages that more closely hew to Standard English employ the use of the word “which,” while broader dialects employ the use of the word “what” for non-personal clauses. (Herrmann 40). “What” is used much more in non-restrictive clauses than other particles, much like the pronoun “who,” which appears in 29.2% of non-restrictive clauses. page 40. “What” appears in 33.5% of non-restrictive clauses, compared to the particles as a whole, which appeared in 19.9% of non-restrictive clauses (page 40). Personal pronoun “who” favors restrictive environments (70.8%), while mainly non-personal pronoun “which,” greatly favors non-restrictive environments (82%), “whom” favors nonrestrictive environments, while “whose” prefers restrictive environments. (Herrmann 41). These basic concepts represent the dialectical variance across the regions studied by Ms. Herrmann. It is with this basic framework established by Ms. Herrmann’s work that I will attempt to analyze the ground-breaking work of Ivan Sag in his article “English Relative Clause Construction,” in the context of the British Isles dialectical patterns. Sag’s work is based on the rubric of the HPSG, which is a type of word hierarchy, with the principles based upon the Head Feature Principle and the Valence Principal. (Sag 6). The head daughter identifies the HEAD value of any headed phrases, indicated on the HEAD values of he lowest VP and all the phrases that contain it. (Sag 6). The Valence Principal requires that “in each phrase the head daughter’s relevant valence feature specify an element that is identified with the appropriate non-head sister of the head daughter.” (Sag 6). MOD is a HEAD feature that guarantees that modified elements are lexically selected by adjuncts. (Sag 6). CONT, which is semantic content, identifies with the adjunct daughter or the head daughter, if there it not an adjunct daughter. (Sag 6). In HPSG, the sign is the fundamental utterance type, and words and phrases are subtypes. (Sag 7). Lexical entries describe words, while schemata describes phrases. (Sag 7). The phrases are classified as either a headed-phrase or a non-headed-phrase, which comes under the broader classification of “headedness.” (Sag 9). The headed phrase is divided into the head-adjunct-phrase and head-nexus-phrase. The head-nexus-phrase is further divided into four subtypes of head-filler-phrase, head-subject-phrase, head-complement-phrase and head-specifier-phrase. (Sag 9). The head-subject-phrase is further broken down into finite head subject phrase. Further, a phrase has a separate branch in same position as “headedness,” known as “clausality.” Clausality is further broken down into four subclasses - imperative clause, declarative clause, interrogative clause, imperative clause and relative clause. (Sag 13). An individual phrase inherits from both clausality and headedness. (Sag 13). The basic structure are subject to constraints, including constraints upon the wh-relative clauses. (Sag 20). The constraints upon the relative clauses headed by a -wh word was described by Sag, who stated that wh-relatives “inherit constraints that require them to be [MC -] and [INV -].” (Sag 21). Another constraint is “that non-head daughter of a wh-relative must have a REL containing exactly one index - the same index as the relative clause’s MOD value.” (Sag 21). The third construct is that the wh-word occurring in the non-head daughter must be coindexed with the nominal that the relative clause modifies.” (Sag 21). The wh- clauses has relatives, the simplest of which is the wh-subj-rel-cl, which is a subtype of both wh-rel-cl and fin-hd-sub-ph, so it has the constraints of both. (Sag 22). Therefore, there is no infinitival instance of this clause. (Sag 23). Sag also examined the usage of the word “that,” stating that, although there is some theory that the word “that” is a pronoun, it is being treated for the sake of Sags constructs as a relative to the wh words, as it is similar to the word “who.” (Sag 31). The behavior of “that,” as compared to the word “who” is virtually identical. Which means that it can be analyzed as a wh-pronoun. (Sag 32). Further, Sag’s article dealt with the adnominal phrases. He divided them into bare phrases, simple infinitival relatives and reduced relatives. The bare relatives are treated as the subtype of non-wh-rel-cl, which inherits from the type fin-hd-subj-ph. (Sag 37). The simple infinitival relatives belong to the subtype non-wh-rel-cl, also is a subtype of hd-comp-ph and the head daughter “must be both infinitival and slashed.” (Sag 37). Another constraint is that the infinitival relatives cannot follow a wh-relative clause and that “sequences of simple infinitival relatives should be possible.” (Sag 38). As for the reduced relatives, the phrase must be specified as both SUBJ PRO and HEAD MOD. (Sag 41). Ms. Herrmann’s article also described a lexical hierarchy, called the accessibility hierarchy. The basic outline of this hierarchy is as follows: subject>direct object>indirect object>oblique case>genitive>object of comparison. (Herrmann 48) The hierarchy constraints were as follows: “1. The language must be able to relativize subjects; 2. Any relative clause forming strategy must apply to a continuous segment of the AH and 3. Strategies that apply at one point of the AH may in principle cease to apply at any lower point.” (Herrmann 48). This hierarchy has been modified over the years. Object of comparison was eliminated, indirect objects are commonly replaced by prepositional objects (eg “he gave her a dog” v. “he gave a dog to her”), while oblique case was renamed PCOMP (prepositional complement) and subject and object complements were grouped with indirect objects and non-prepositional adverbials under the broad category “OTHER.” (Herrmann 50-51). So, the modified version of the AH is as follows: SUBJ>DO>PCOMP>GEN, with PCOMP being divided into prepositional objects (OP), prepositional complements functioning as adverbials (A) and prepositional complements functioning as genitives (G). The other broad category is OTHER, with the sub-categories of subject complements, object complements, indirect objects and nonprepositional adverbs. (Herrmann 51). Herrmann found the Accessibility Hierarchy in all six regions studied, with the relativization being by far the most frequent (63.2%), followed by direct objects at 27.2% and prepositional complements at 7.3%. (Hermmann 51). According to Sag, “relative clauses are all projected from empty relativizers.” (Sag 42). “Who” is the only word that can “relativize all positions on the Accessibility Hierarchy including GEN.” (Herrmann 57). “Which” is also well-represented in all the positions within the modified hierarchy, except GEN. (Herrmann 57). Relative particles “that” and “zero” are in both the lower and higher positions, as are -wh pronouns. (Herrmann 57). Relative particles by far dominated the landscape, with the Central Midlands having twice the number of relative particles as wh-pronouns, and Northern Ireland having 40 times as many. (Herrmann 57). Direct objects and prepositional complements have more phrases featuring “that” and “zero” than wh-pronouns, while the word “that” is prominent in the SUB position. (Herrmann 57). Meanwhile, the words “what” is on the rise, while “as” is declining. “What” seems to start at the subject position and if the words “what” is used as a genitive pronoun, it is also used as a object pronoun and as a subject pronoun, according to the data pulled by the Central Midlands. (Hermmann 58). While the word “what” is currently introduced as a subject pronoun, this is different than its genesis. In Middle English, the word “what” entered the hierarchy at the low position, and mainly in formal and written language. (Herrmann 58). The Central Southwest has an even higher incidence of “what,” as it is “second strongest in the lower positions DO and PCOMP, as well as SUBJ.” (Herrmann 58). On the other hand, East Anglia, where “what” originated, it is dying out, as it has a reputation there as being a vestige of old and vulgar language, which makes it unpopular. (Herrmann 58). “As” is suffering a similar fate in the Central Southwest and East Anglia, retreating to the higher positions of the AH, and will probably soon be obsolete, exiting via the subject position. (Herrmann 58). The only area where “as” occupies a lower position on the AH is the Central Midlands, where it is in the PCOMP position. (Herrmann 59) The data indicates that “what” is driving out “as.” (Herrmann 59). The word “which” is also examined at length by Ms. Herrmann, with her studies finding that the which examples were all non-restrictive clauses, finding that it is “the dominant relative marker in nonpersonal nonrestrictives and a viable option for personal nonrestrictives, whereas the relative markers who, that and zero are only minor options in nonrestrictive relative clauses.” (Herrmann 77). It can also function as a connector, as opposed to a relative pronoun, where its simple function is to connect two statements together. (Herrmann 87). This analysis receives support “when the subsequent clause shows no grammatical gap, ie, when which does not seem to have a grammatical function in the clause, as all syntactic positions are already taken by other noun phrases.” (Herrmann 87). However, Herrmann posits that if you look closer at the connector occurrences of the word which, it seems that it isnt a connector at all, but, rather, one of three distinct constructs – either a preposition ellision, a resumptive or a new start. (Herrmann 88). Prepositional elision is a common occurrence in the English language, especially in the spoken language (Herrmann 89-90). It occurs with different relative markers and relative clauses, such as infinitival clauses and frequently found after “massive or complex intervening material.” (Herrmann 90). In this case, which cannot function as a prepositional complement, instead functioning as an adverbial or indirect object. (Herrmann 90). Resumptive pronouns have a redundant nature, therefore, the “respective grammatical function is filled twice in a relative clause: once by the relative marker and a second time by the resumptive.” (Herrmann 91). A resumptive is found with all kinds of relative markers. (Herrmann 92). New starts are the third type of occurrence associated with the word “which.” (Herrmann 92). These comprise abandoned relative clauses, using a paratactic construction. (Herrmann 93). One of the other constructs described by Sag article is pied-piping, described as “a matter of the inheritance of nonempty specifications for the features REL or QUE, triggered by the presence of a wh-relative or wh-interrogative word within a given phrase.” (Sag 17). In the Sag construct, words are distinguished by whether or not their specifications are non-empty for the features REL and QUE, with the relative and interrogative words bearing this special distinction. (Sag 18). The wh- words pass up non-empty REL and QUE specifications to the phrases that directly dominate them, so successively larger phrases. (Sag 18). This results in the requirement that the “non-head daughter have a non-empty REL or QUE specification and this will be sufficient to guarantee that the non-head daughter contains an appropriate wh- word somewhere within it.” This is known as “feature passing,” and it is facilitated by words. (Sag 18). Words project their REL specifications upwards via their dependents. (Sag 18). Pied-piping is related to the British Isles dialect by Ms. Herrmann, by describing it as one of two basic kinds of prepositional phrasing, the pied-piping and the preposition stranding. Ms. Herrmann describes the concept of Pied-piping as preposition fronting, where the preposition is moved to its initial position with the wh-marker and governs the relative pronoun. (Herrmann 45). An example of pied-piping is as follows: “I was drinking a glass of wine, for which I paid $6.” This construct is typical of written Standard English. In the use of dialects, however, another type of prepositional phrasing, known as preposition stranding is typically employed. This is where the “preposition is left behind in its normal clause position without its complement.” (Herrmann 45). An example of this is “my computer, that I used to write my paper with, is a HP Pavilion.” (Herrmann 45). Once again, the dialectical employ of either the prepositional stranding or the pied-piping indicates whether the dialect of a sub-set more closely hews to Standard English than other sub-sets. Ms. Herrmann notes that pied-piping is more common in written Standard English, yet the dialects vastly preferred the stranding method, with 93.5% of the studied dialects preferring this method. (Herrmann 46). She attributes this partially to the fact that only relative pronouns allow both kinds of phrasing, while relative particles only allow the stranding. (Herrmann 46). She discovered that in every instance where the marker was that, what, as or zero, the prepositions were stranded. (Herrmann 46). Meanwhile, even in cases of the relative pronouns, only 31.6% were examples of pied-piping, while the prepositional stranding made up the rest of the 68.4%. (Herrmann 46). What this showed is that the prepositional stranding was preferred even when the pied-piping was permitted. (Herrmann 46). In conclusion, while Sags theory is novel and ground-breaking in the field of linguistics, Herrmann shows that at least some of the novel theory can be applied to regional dialects of the British Isles. Since the theory that Sag propounds is designed to fit any kind of language construct, this would seem to be self-evident. Herrmanns work is important, as it shows the characteristics of the regional dialects that are closer to Standard English than other dialects and it also demonstrates the commonalities between the regions are far as their grammatical structure. While Sags theory is more broad, but can be narrowly tailored to fit Herrmanns, or any other language construct. Works Cited Sag, Ivan A. “English Relative Clause Constructions.” Journal of Linguistics May 1997: 1-51. Herrmann, Tanja. “Relative Clauses in English Dialects of the British Isles.” 21-111. Read More