Hypocorrection

Originally, hypocorrection, or an accented pronunciation of words, may have stemmed from physical properties involved in sound production, such as aeroacoustics, anatomy and vocal tract shape.^[3]

Hypocorrection may also result from the failure to cancel coarticulatory effects. Ohala mentions that hypocorrection happens when a listener fails to make use of compensation or, to be exact, when the listener lacks experience with a series of contextual discrepancies that allows them to execute such correction or cannot detect the conditioning environment for various reasons such as noise and the filtering associated with communication channels.^[4]

When a listener restores a phoneme from its contextually-influenced realisation, normal speech perception involves the process of correction. That is in accordance to a model proposed by John Ohala which involves synchronic unintended variation, hypocorrection, and hypercorrection. For example, in a language that contains no contrasting nasality for vowels, the utterance [kɑ̃n] can be reconstructed, that is, "corrected" by the listener as the phoneme sequence [kɑn] that was intended by the speaker because they have the knowledge that every vowel is nasalized before a nasal consonant. Hypocorrection occurs if the speaker fails to restore a phoneme, perhaps because the [n] was not pronounced very clearly, and analyses the utterance as [kɑ̃].^[5]

However, further studies suggest that there could be another possible reason for the occurrence of hypocorrection: variation in the compensation. For example, Beddor and Krakow (1999) tested American listeners' nasality judgments on the nasalised vowel [Ť]/[õ] between nasal consonants ([m ڧ ԝn]), on oral vowels [Ť]/[o] between oral consonants ([bVd]), and on the same oral vowels in isolation ([#V#]). They found that 25% of [ԝ] in nasal contexts were heard as more nasal than [V] in oral contexts, which shows that compensation was incomplete or irregular. In addition, Harrington et al. (2008) illustrated systematic variation in compensation between young and old listeners. They contrasted the two groups' identification of a vowel from an /i/-to-/u/ continuum in palatal ([j_st]) and labial ([sw_p]) contexts. Both groups' category boundaries were at comparable points on the palatal continuum and were closer to the /i/-end than on the labial continuum, which shows a compensation effect. However, the younger group's boundary on the labial continuum was much closer to the boundary on the palatal continuum, which demonstrates less compensation in comparison to the older group. These results indicated a difference in the listeners' own speech production: the /u/ for younger speakers was more fronted than that of the older speakers in general. The findings indicated that listeners compensate for only as much coarticulation as is expected in their own grammar, and that form of "grammar" is affected by the listener's previous linguistic experience. That could thus add to Ohala's list of causes of hypocorrection differences in the coarticulation/compensation norm between a speaker and a listener, which could result in events by which a listener uses compensation and still fails to extract from a heavily-coarticulated speech segment "the same pronunciation target intended by the speaker."^[4]

As for the realm of the social aspect, the intentional use of hypocorrection or, for example, affecting a Southeastern American accent to sound less elitist involves "make-believe hesitations and colloquial language" that "work as affiliative strategies (softeners) etc."^[6] Over time, hypocorrection has emerged by both physical features of voice production and affected accents, and it is typically used by people who do not wish to associate themselves with overly-sophisticated local dialects. Hypocorrection also works as a softener.^[7] Some forms of hypocorrection are attempts to give one's discourse a clumsy, colloquial, or even a broken and dysfluent style in introducing clever or innovating statements or ideas. More often than not, hypocorrection allows the speaker, by toning down a potential self-flattering image, to avoid sounding pretentious or pedantic, thus reducing the risk of threat to the recipients' faces. That can be linked to the politeness theory, which accounts for politeness in terms of the "redressing of affronts" to a person's sociological face by face-threatening acts.^[8] The theory elaborates on the concept of face (to "save" face or to "lose" face) and discusses politeness as a response to alleviate or avoid face-threatening acts that include insults, requests etc. Therefore, hypocorrection may be used in such situations to allow people to save face.

Hypocorrection may have a part in innovating sound change. Ohala proposed a theory of sound change arising from the listener's misperception.^[9][10] The theory highlights important variations in "the phonetic form of functionally equivalent speech units" and puts forth that when faced with coarticulatory speech variation, listeners do one of the following:

• They perceptually compensate for predictable variations and arrive at the pronunciation target intended by the speaker.
• they fail to compensate for coarticulation and assume that the coarticulated form is the intended pronunciation.

The first situation describes what happens in normal speech perception and the second situation describes what happens in hypocorrection, which is the type of misperception in the perceptual compensation for /u/-fronting. Hypocorrection is the underlying mechanism for many assimilatory sound changes, and the main concept of hypocorrection is that contextually-induced perturbation is regarded by a listener as a deliberate feature of the speech sound. Hence, hypocorrection has the potential to change the listener's phonological grammar by what Hyman called "phonologisation," a process by which intrinsic or automatic variation becomes extrinsic or controlled.^[11] For years, many researchers have analysed sound change as a result of phonologisation^{[12][13][14][15]} which underscored the theoretical significance of hypocorrection as a condition for sound change via phonologisation.^[4]

The listener misperception hypothesis of sound change^[16][17][18] has been a worthwhile domain of inquiry over the years, partly because it makes testable predictions. According to the area of research, phonological rules arise by mechanical or physical constraints inherent to speech production and perception. The perceptions involve the likes of listener hypocorrection and hypercorrection. Cross-linguistic tendencies in grammars are therefore thought of as "the phonologization of inherent, universal phonetic biases".^[19] Hypocorrection is formally symmetrical and so there is no basis for the unidirectionality of sound changes. For example, consonants normally palatalize, rather than depalatalize, before front vowels, which has no inherent explanation. That ambiguity begs for reanalysis, but something else must demonstrate the directionality of the change. Assimilation and dissimilation are quite different in other ways as well since dissimilation (by hypothesis, hypercorrection) never gives rise to new phonemes, unlike assimilation (via hypocorrection). Such inherent asymmetries are not predicted by the theory as it stands.^[20]