Introduction
This paper focuses on the standard Beijing dialect since Mandarin is spoken in different regions with varying accents. Studying potential schematic representations of the dialect’s phonological processing is a major field of research. In this case, conducting a phonological association examination of the language is critical to understanding the development of its standard components relative to other languages.
Nonetheless, this critical analysis will rely on optimal theory (OT) to explore the modification of Mandarin words to understand the dialect’s illicit codas and consonant clusters. The theory will provide a framework for a multidimensional review of Mandarin phonemes’ characteristics. The paper will argue that some Mandarin words are borrowed from American English. It will seek to illustrate that no hypothetical or complex rules are involved in the modification of Mandarin loanwords. However, simple deletion, epenthesis, or feature change could be used to construct them. It will demonstrate that well-formedness constraints of the Mandarin linguistic elements trigger these repair strategies for its syllable structures.
Syllable Structure and Repair Strategies
According to the syllable structure, the following are observable. (C)(G)V(N) is the Mandarin syllable structure, where G = glide and N =nasal (Guo, 1999). While coda, prenuclear glide, and onset are optional, the nucleus is obligatory. Syllable types in Mandarin are: V, GV, VN, GVN, CV, CGV, CGVN. However, apart from CG-combination, the Mandarin syllables lack consonant clusters. On the other hand, (C)(C)(C)V(C)(C)(C) is the syllable structure in English.
These structures directly affect the initial rules by the employment of which English American loanwords behave in Chinese Mandarin. Three main repair strategies—featural change, epenthesis, and deletion- serve as protocols for English words adoption into Mandarin. (Guo, 1999). Both onset and coda clusters in consonant structures change their form when adopted by Chinese speakers, and it is of particular interest to find out the means of how they operate.
There are not many tri-consonantal clusters that abide by the rules of Mandarin phonology; however, a multitude of examples can be found when dealing with di-consonantal sounds. As for the di-consonantal clusters, Mandarin mostly adopts two strategies to modify them—insert vowels to syllabify the consonant(s) or delete the consonant(s), as (6) illustrates (Guo, 1999).
To meet Chinese Mandarin phonetic standards, the consonant clusters are met with an additional vowel, albeit some syllables are skipped. Examples 6(a) and 6(b) indicate how well onset and coda clusters are transferred from English to Mandarin “with the insertion of vowels to satisfy the syllable structure constraint” (Guo, 1999). However, this strategy of transferring consonantal syllables is not the only one used.
The practice of a word acquiring new sounds is epenthesis – it is another strategy characteristic of foreign English words in Chinese Mandarin phonology. Another is omitting consonantal clusters to fit the scheme of Mandarin s syllable structure, which differs from English radically. Examples (7b) and (7c) represent the prohibited codas in Mandarin. Different strategies are employed: (7b) inserts vowels after the illegal codas [k] and [s] to syllabify them, whereas (7c) deletes the banned coda [1] (Guo, 1999). These strategies are exemplified further.
Rule-Based Analysis and Its Limitations
Some linguistic scholars propose that rule-based analysis could be used to examine the phonological structure of Mandarin words borrowed from the American English dictionary. Several insertions and deletion rules should be designed to account for how a person could modify the syllable-initial or –final English consonant clusters. Chang claims that the account should further illustrate how the modified clusters would conform to Mandarin syllable structures including the insertion of nasals or stops [i] or [o] between Consonant + liquid and [i] between syllable-initials [s], and deleting word-final or postvocalic [r] consonants (as cited in Guo, 1999, p. 196). However, the rule-based analysis faces various problems or limitations thus making it unsuitable for doing phonological analysis.
Rule-Based Limitations
First, the origin of deletion and epenthesis rules is ambiguous. The Standard Theory demands that such rules originate from English or exist in Mandarin (Berces & Honeybone, 2020). Nonetheless, as Guo (1999) claims, there is no evidence about epenthesis rules that could allow vowels after illicit codas and C-cluster consonants. Therefore, given that no epenthesis rules in both English and Mandarin cast aspersions about the possibility of using rule-based analysis in phonological studies.
Second, some members of the identical context within various phonemes databases cannot be modified by some rules. For instance, sometimes, syllable-final [I] could be deleted, but others are parsed (Guo, 1999). However, no rule explains why in some cases the syllable-final is parsed (Berces & Honeybone, 2020). According to Guo (1999), the abovementioned issues could be solved through the optimality theory approach. He further notes that grammar comprises common ranked constraints acceptable across the globe. But, more importantly, this concept does not follow any universal rule.
Optimality Theory in Mandarin Phonology of Loanwords
Optimality Theory (OT) has tremendous implications on how people analyze a linguistic phenomenon, their perspectives on language, and how it relates holistically to human cognition (van de Weijer, 2019). It is well suited in conflicting areas such as between phonetic pronounceability and morphemic structure, or between syllable sound and its structure. Van de Weijer (2019) explains that this suitability is integrated into OT’s architecture, designed purposely for conflict resolution. OT assumes “that the grammar of languages consists of a set of ranked violable well-formedness constraints” (Guo, 1999, p. 196). Although these constraints are universal, their ranking is specific to a given language.
Input and output correspondence is the main focus of the optimality theory. But, although in “standard theory the output is derived from an input via a context-driven rewrite rule, in OT the output is chosen from a set of candidates which are associated with an input” (Guo, 1999, p. 196). Based on this view, OT follows a generative rule in which the “generative” idea is implied in its “sense of derivational” (van de Weijer, 2019, p. 120). It has two levels: the lexical or abstract level (consists of the inputs) roughly considered as the underlying forms, which are modified into output or the phonetic or surface forms.
OT’s basic structure is not innate, as any linguist would believe. However, it arises from general cognitive strategies that shape people’s knowledge and ability to process languages (van de Weijer, 2019). Perhaps OT is not specific to any language.
Retrospectively, OT does not have specific rules because the admission of a candidate’s analyses as assessed by the constraint hierarchy is determined by the common characteristics of well-formedness. Parallel computation is used to identify a candidate’s best satisfaction or fatal violation thus eliminating the need for derivational processes. Furthermore, the central focus of OT language analysis is the output or surface structure (Guo, 1999). This notion makes OT a perfect fit for describing loanword phonology. Mandarin, as the loan language may introduce underlying forms, which are not stimulated by itself. However, given that the underlying forms are ultimately forced to align with the output constraints in the borrowing language (Mandarin), the foreign underlying representations arise as the output patterns of the loan language, in this case, Mandarin. Thus, by just inspecting the outputs or the surface representations, the constraints, including their ranking or dominancy in a certain language could be derived.
Representation-Based Approach
Another phonological model that can be utilized to analyze Mandarin Chinese phonology tendencies from American English loanwords is Representation-based Phonology. The theory can be summarized by the following, “sound patterns and processes can be described, and their workings (including triggers, targets, and outputs) explained, with reference to a hierarchical structure that can be represented by tree diagrams” (Berces & Honeybone, 2020). Currently, this approach classifies as one of four main phonology analysis traditions, and has a number of strands within it as well. It is also “highly relevant that phonology, in general, has two major aspects: processes/computation and representations” (Berces & Honeybone, 2020). This way, the representation approach is an established part of a phonological analytical process, that can be used in pairs or separately from other forms.
In support of this idea, Berces notes that (2020) “computation and representations are two separate modules of phonology and as a consequence, any theory of representation should be freely combinable with any theory of computation finds further support.” These theories were adapted both in the practices of computation and representation. He continues by giving an example of this, the contrastive hierarchies of the Toronto School are often coupled with RBP; LR has also been applied in GP” (Harris, as cited in Berces & Honeybone, 2020). The relevance of the theory may also be rooted in the fact that “many forms of suprasegmental phonology including metrical phonology, intonational phonology, and tonal phonology are fundamentally representation-based in nature” (Berces & Honeybone, 2020). It is clear that the representative-based approach is used widely in various forms of hybrid models of phonological analysis.
There are other instances of usage and application of this model. Berces (2020) mentions the nature of “pure” representation methods as well, “both the requirement of the local source (in GP, the Non-Arbitrariness Principle) and the rejection of extrinsic rule ordering (in GP, due to the Minimality Hypothesis). The author (2020) explains their origin lies in “the assumption that our model of phonological knowledge should be constrained enough to avoid overgeneration and make predictions that are falsifiable.” Also, the notion of the local source is central to RepBP’s classification of phonological phenomena into two basic types (and it is ruling out other types, like RBP’s simple ‘feature-changing rules’) – assimilations and lenitions (Berces, 2020). On this base, Mandarin Chinese can be analyzed as a case of either of the two. The author paints an accurate picture of representation-based analysis, which can be employed for the particular case of this research, either coupled with CBP or alone.
Subsegmental structure
As has been mentioned prior, there is a number of subordinate theories in Representative-based Phonology. With one of them, Segment-internal, melodic representations RepBP share a number of general properties (Berces & Honeybone, 2020). In particular, as Harris state, they accept the ‘one-mouth’ principle (as cited in Berces & Honeybone, 2020, p. 118), i.e., they employ the same set of melodic primes for the representation of consonants and vowels, or at least there is considerable overlap between the two sets. (Berces & Honeybone, 2020). Another two, are comprised of a principle of direct phonetic interpretation in both simple and complex phonological expressions, and thirdly, segment-internal dependency relations. These concepts are key between the representative approach and subsegmental structure.
Constraint-Based Analysis
Guo (1999) highlights the Basic Syllable Structure Constraints, as suggested in the 1993 Prince and Smolensky’s phonological review. Accordingly, the CV-combination ranks as the most unmarked structure of syllables. Prince and Smolensky, as cited in Guo (1999), note:
- ONSET: Syllables must have onsets.
- NOCODA: Syllables must not have a coda.
- Nuc (nucleus): Syllables must have nuclei. 13
- *COMPLEX (no complex): No consonant cluster is allowed within a syllable. (where consonant cluster means [+cons] [+cons]…)
Mandarin allows codas, which are highly restricted to the form [n] or [ᵑ] (p. 197). Therefore, NACODA might be modified into CODACON, as illustrated in (12), which:
CODACON: Syllables must have no coda, apart from a velar nasal or alveolar
Besides, OT demands that consistency should be met. That is, input and output should have a “personalized” correspondence (Guo, 1999).
- MAX-I0 (maximum-input/output, first version)
Every input segment must have a corresponding segment in the output.
- DEP-IO (dependent-input/output)
Every output segment must have a corresponding segment in the input.
MAX-10 requires no deletion because any segment deleted in the output will cause some input segments to lack correspondents in the output. On the other hand, DEP-10 does not crave insertion because any segment included through later insertion in the output would not have an equivalent input segment.
Relative to the data, in the absence of deletion, epenthesis is triggered to circumvent C-clusters. This aspect satisfies:
*COMPLEX, constraint MAX-10 will not be violated, but DEP-10 will be violated in Mandarin. Accordingly, this premise means on the ranks, *COMPLEX is higher, followed by MAX-10, then DEP-10. Segments’ deletion would be common if MAX-10 ranked lower than DEP-10. Equally important is the ranking among DEP-10, MAX-10, and CODACON. Shunning illicit coda with an exception of the liquid codas triggers insertion strategy. This occurrence implies that DEP-10 is the least ranked, followed by MAX-10, the CODACON is predominant. Table (7) demonstrates this interaction among the constraints.
While *COMPLEX and CODACON are ranked, the ranking between them is insignificant in identifying the optimal candidate. Illustratively, the two constraints do not interact. The following schema shows this ranking between the two (11):
*COMPLEX, CODACON>>MAX-10>>DEP-10
The ONSET status here means that the penalty for its violation is low because the onset less syllables are numerous in Mandarin language loanwords and forms. However, constraint ONSET could be violated to retain faithfulness as illustrated in Tables 8 and 9.
Tableaux 8 and 9 emphasize the note that the output constraints DEP-10 and MAX-10 are higher in rank than ONSET in Mandarin loanwords and native forms. However, the Alabama illustration presents a counterexample. In the Alabama case, the ONSET ranks higher than DEP-10. However, as Guo (1999) opines, such counterexamples where the ONSET is dominant compared to the constraint DEP-10, are rare. He further states that other factors including character choosing as well as semantic-ambiguity avoiding could interact to modify such loanwords into Mandarin.
Thus, the general ranking of the constraint could be written as (13).
*COMPLEX, CODACON>>MAX-10>>DEP-10>>ONSET
The constraint interaction as shown in the above illustration (13) elaborates many cases in which English codas and consonant clusters are repaired or modified in Mandarin while at the same time leaving the liquid codas’ behavior unresolved. Remarkably, the general constraint ranking envisages why liquid codas are parsed everywhere though the data display shows otherwise, that, the lack of salience in polysyllabic words allows liquid codas to be unparsed. Yip, as cited in Guo (1999), states that “unsalient segments could be faintly visible, and thus may be overlooked by the MAX-10 constraint” (p. 199).
Slight modifications as indicated in (14) depict the MAX-10’s final statement.
(18) MAX-10 (final version): “Every (salient) of the input has a correspondent in the output,” (Guo, 1999, p. 199). In this case, the unsalient segments are usually overlooked and therefore unparsed. However, they are parsed sometimes. A key question about when the parsing of the unsalient liquids occurs. Responding to this question, Guo (1999) expounded that when a coda follows a liquid coda, for instance, Mort and Bart, the liquid is unparsed. But while in monosyllabic words, for example, Neil, Gil, and Dale, “the liquid coda is parsed” (Guo, 1999, p. 199).
Conclusion
Following the American English words, this study examined the constraint-based analysis of loanwords in Mandarin phonology. It relied on the optimality theory to reveal the underlying forms and the modification of the English input into Mandarin output. The critical review revealed that there is no need for complex rules in modifying English words into Mandarin representations. According to the optimality theory, this change occurs through simple deletion, epenthesis, or feature change. Unlike rule-based analysis, OT does not have specific documentation. The admission of a person’s evaluations, as assessed by the constraint hierarchy, is influenced by the universal characteristics of well-formedness. OT relies on parallel computation to identify a candidate’s best satisfaction or fatal violation. In this case, it eliminates the need for derivational processes. From this paper analysis, OT is a perfect fit for describing Mandarin loanword phonology. Mandarin, as the loan language may introduce underlying forms, which are not self-inspired. However, given that the underlying forms must align with the output constraints in the borrowing language (Mandarin), the foreign underlying representations arise as the output patterns of the loan language, Mandarin.
References
Berces, K. B., & Honeybone, P. (2020). Representation-based models in the current landscape of phonological theory. Acta Linguistica Academica, 67(1), 3-27.
Guo, H. L. (1999). Mandarin loanword phonology and optimality theory: Evidence from transliterated American state names and typhoon names.[Paper presentation]. Proceedings of the 13th Pacific Asia Conference on Language, Information and Computation, 1999, Taipei, Taiwan. Web.
van de Weijer, J. (2019). Where now with optimality theory?Acta Linguistica Academica, 66(1), 115-135. Web.