On “Whistle” Sound Objects in English Everyday Conversation

ABSTRACT

In this article we study the forms and functions of whistling in social interaction. Our analysis identifies two basic forms of conversational whistling, (a) melodic whistling, when participants whistle the tune of, e.g., a familiar song; and (b) nonmelodic whistling. The focus in this article lies on nonmelodic whistles, which come in two contours linked to specific actions: (a) the tonal whistle deployed for summoning (e.g., a domestic animal but also human participants); and (b) the gliding whistle used for affect-laden responses to informings that breach a norm, often ones containing a numerical reference. The pitch contour used on the latter type of whistle matches those found for more lexical sound objects, e.g., oh, ah, and wow. The data base for the study comprises a wide range of audio and video recordings of mundane American and British English telephone and face-to-face conversations.

Notes

¹ In the conversational data we have examined, the whistles used do not exceed 2500 Hz in fundamental frequency (see the following).

² Sound objects are defined as “conversational objects with [phonetic substance but] minimal semantic content.” The term sound object “aims to reflect the fact that these objects are spoken language resources for which the sound pattern and its context-specific use are distinctive for the meaning.” Included are “so-called ‘primary interjections,’ such as, e.g., oh, ah and ooh, and non-lexical sounds such as clicks and whistling, which have been found to function similarly in talk-in-interaction” (Reber, 2012, p. 12; see also, 2009, Fn 5).

³ In linguistics, there is a general distinction between primary interjections (e.g., oh, ah or ugh), and secondary interjections (e.g., oh my God, good heavens). This distinction originated with work by the German psychologist Wilhelm Wundt at the beginning of the 20th century, in which he describes primary interjections as remnants of non-linguistic Naturlaute (natural sounds) that interrupt speech for the expression of emotion. On the other hand, secondary interjections are treated as linguistic forms of pure emotional expression (Wundt, 1911, pp. 320–321).

⁴ See also Dingemanse, Torreira, and Enfield (2013), who argue that huh has the status of a universal word.

⁵ The data used for our analysis were collected with informed consent, and names, images, and other identifiers have been anonymized. In the case of the NB corpus, which was compiled in the 1960s, the origins are unknown.

⁶ Our thanks to Giovanni Rossi for sharing this data extract with us.

⁷ Presumably cat-calls would belong in this category, although in addition to summoning, they incorporate an element of appreciation or aggression as well. We do not have any examples of cat-calls in our collection.

⁸ Our thanks to Elliott Hoey for providing us with this example.

⁹ This whistle has an offglide but we categorize it as “tonal” because it is sharp and fast in contrast to “gliding” whistles, which are smooth and slow.

¹⁰ Our thanks to Barbara Fox for providing us with access to this example.

¹¹ The phrase rUn part of that is produced with creaky voice.

¹² It may be worth noting that earlier in the conversation Gary has explained to Julie that he builds barns for a living: He is thus an expert in this domain.

¹³ The span covered by the gliding whistles in our collection ranges from 4 to 12 semitones.

¹⁴ The “whispery” quality presumably comes from the whistle being softer with more friction at the lips (personal communication, Richard Ogden).

¹⁵ It is true that the IPA has a phonetic representation ([w͎]) for “whistle,” but this representation does not necessarily correspond to the sounds observed in our data nor does it capture the difference between a “tonal” whistle and a “gliding” whistle.

¹⁶ Historically, crumbs seems to stem from the mild oath By crum! (https://www.merriam-webster.com/words-at-play/edible-interjections/crumbs).