2.1Procedural encoding

Besides the presumption of optimal relevance, speakers can guide utterance interpretation more specifically using linguistic expressions that explicitly encode “information about the inferential process that the addressee should use” (Blakemore 2002Blakemore, Diane 2002 Relevance and Linguistic Meaning. Cambridge: Cambridge University Press. Blakemore, Diane 2002 Relevance and Linguistic Meaning. Cambridge: Cambridge University Press. , 90). Blakemore (2002Blakemore, Diane 2002 Relevance and Linguistic Meaning. Cambridge: Cambridge University Press. Blakemore, Diane 2002 Relevance and Linguistic Meaning. Cambridge: Cambridge University Press. , 78) describes such linguistic expressions as encoding procedural information, distinguishing it from conceptual information, which are propositional representations of the world that are encoded by most lexical items, e.g. big domestic cat in (7a) above. For example, the relationship between the two statements (9a) and (9b) can be interpreted in two ways, (10a) and (10b), depending on whether the discourse connective so or after all is used:

Blakemore (2002)Blakemore, Diane 2002 Relevance and Linguistic Meaning. Cambridge: Cambridge University Press. Blakemore, Diane 2002 Relevance and Linguistic Meaning. Cambridge: Cambridge University Press. argues that so in (10a) instructs the addressee to combine (9a) (repeated as [11a]) and the existing assumption (11b) that has been made salient in his cognitive environment to derive the contextual implication (11c), which acts to reinforce the utterance (9b):

Conversely, after all in (10b) instructs the addressee to combine (9b) (repeated as [12a]) and the existing assumption (12b) made salient in his environment to derive the contextual implication (12c), which acts to reinforce the original utterance (9a):

Under the relevance-theoretic framework, particles and other discourse connectives have been analysed as encoding similar procedural instructions (e.g. Jucker 1993Jucker, Andreas H. 1993 “The Discourse Marker Well: A Relevance-Theoretical Account.” Journal of Pragmatics 19: 435–452. Jucker, Andreas H. 1993 “The Discourse Marker Well: A Relevance-Theoretical Account.” Journal of Pragmatics 191: 435–452. on English well; Ler 2005Ler, Vivien 2005 “An In-Depth Study of Discourse Particles in Singapore English.” PhD dissertation. National University of Singapore.Ler, Vivien 2005 “An In-Depth Study of Discourse Particles in Singapore English.” PhD dissertation. National University of Singapore. on the CSE particles meh, hor, and lah; Ljungqvist 2007Ljungqvist, Marita 2007 “ Le, Guo and Zhe in Mandarin Chinese: A Relevance-Theoretic Account.” Journal of East Asian Linguistics 16: 193–235. Ljungqvist, Marita 2007 “Le, Guo and Zhe in Mandarin Chinese: A Relevance-Theoretic Account.” Journal of East Asian Linguistics 161: 193–235. on Mandarin le, guo, and zhe; Schourup 2011Schourup, Lawrence 2011 “The Discourse Marker Now: A Relevance-Theoretic Approach.” Journal of Pragmatics 43: 2110–2129. Schourup, Lawrence 2011 “The Discourse Marker Now: A Relevance-Theoretic Approach.” Journal of Pragmatics 431: 2110–2129. on English now), and this paper will likewise analyse the particle hor as encoding a procedural instruction that guides the addressee’s pragmatic interpretation of the hor-marked utterance.

2.2Interrogatives and directives in relevance theory

Since hor-marked utterances do not just function as assertions but also as questions (e.g. [1]) and directives (e.g. [3]), it is important to understand how non-declarative utterances are analysed in relevance theory. According to Sperber and Wilson (1988)Sperber, Dan, and Deirdre Wilson 1988 “Mood and the Analysis of Non-Declarative Sentences.” In Human Agency: Language, Duty and Value, ed. by Jonathan Dancy, J. M. E. Moravczik, and C. C. W. Taylor, 77–101. Stanford: Stanford University Press.Sperber, Dan, and Deirdre Wilson 1988 “Mood and the Analysis of Non-Declarative Sentences.” In Human Agency: Language, Duty and Value, ed. by Jonathan Dancy, J. M. E. Moravczik, and C. C. W. Taylor, 77–101. Stanford: Stanford University Press., non-declarative sentences encode representations of potentiality and desirability – imperative sentences represent descriptions of states of affairs that are potential and desirable, while interrogative sentences represent descriptions of thoughts that are desirable.

For imperative sentences, the illocutionary force of the utterance is also determined by whether the potential state of affairs described is desirable to either the speaker or the addressee – a directive encodes a proposition describing a potential situation that is desirable to the speaker, while advice, instructions, and the granting of permission describe potential situations that are desirable to the addressee (Sperber and Wilson 1988Sperber, Dan, and Deirdre Wilson 1988 “Mood and the Analysis of Non-Declarative Sentences.” In Human Agency: Language, Duty and Value, ed. by Jonathan Dancy, J. M. E. Moravczik, and C. C. W. Taylor, 77–101. Stanford: Stanford University Press.Sperber, Dan, and Deirdre Wilson 1988 “Mood and the Analysis of Non-Declarative Sentences.” In Human Agency: Language, Duty and Value, ed. by Jonathan Dancy, J. M. E. Moravczik, and C. C. W. Taylor, 77–101. Stanford: Stanford University Press.; Clark 1991Clark, Billy 1991 “Relevance Theory and the Semantics of Non-Declarative Sentences.” PhD dissertation. University College London.Clark, Billy 1991 “Relevance Theory and the Semantics of Non-Declarative Sentences.” PhD dissertation. University College London.). In this paper, I will focus only on directives as hor is not used with other types of imperative sentences. The authoritative force of a directive results from evaluating the proposition described by the directive based on other contextual assumptions that are held by the addressee. For instance, the directive “Give me a drink!” might be processed by the addressee roughly as follows:

Since the linguistic meaning of the directive only conveys the achievability and desirability of the state of affairs described, a declarative that conveys the same meaning can also be interpreted as a directive as it is processed further by the addressee in order to derive optimal relevance, e.g. (14):

Based on the presumption of optimal relevance, B would interpret A’s utterance as indicating, not just her need for a drink, but that it is potential and desirable that she be given a drink (since this leads to greater cognitive effect by deriving a contextual implication). This interpretation will thus lead B down a similar inferential route as with (13) even though the utterance is in the form of a declarative.

In contrast to directives, questions are taken to be representations of thoughts desirable to the speaker. Sperber and Wilson (1988Sperber, Dan, and Deirdre Wilson 1988 “Mood and the Analysis of Non-Declarative Sentences.” In Human Agency: Language, Duty and Value, ed. by Jonathan Dancy, J. M. E. Moravczik, and C. C. W. Taylor, 77–101. Stanford: Stanford University Press.Sperber, Dan, and Deirdre Wilson 1988 “Mood and the Analysis of Non-Declarative Sentences.” In Human Agency: Language, Duty and Value, ed. by Jonathan Dancy, J. M. E. Moravczik, and C. C. W. Taylor, 77–101. Stanford: Stanford University Press., 23) propose that “a thought is desirable only if it is relevant”, and analyse polar questions and wh-questions as follows:

The presumption of optimal relevance also guides how the speaker interprets answers. In cases where the answer supplied does not match the desired thought that is encoded by the question, the speaker asking the question will still assume that the response is optimally relevant and try to derive the immediately relevant proposition from it, e.g. in (17):

Although A did not respond with a proposition that is immediately relevant to B, B assumes that A’s answer is still intended to be optimally relevant, and so derives the relevant answer (17c) by interpreting the stated utterance (17a) together with a contextual assumption (17b) that is assumed to be part of the common ground between speaker and addressee.77.While this is the best-case scenario, it is also possible for A to be mistaken in what she believes to be common ground, e.g. if B has a low opinion of MIT graduates and believes that A shares his view, or for B to be unable to recognise (17b) as relevant for one reason or another, which would lead to a misunderstanding or communicative failure in the conversation. In this case, the additional processing costs are offset by the additional cognitive benefit that A intends, i.e. the implication that it is obvious that John is smart given that he graduated from MIT.

2.3Mood and intonation contour

Halliday (1970)Halliday, Michael A. K. 1970 A Course in Spoken English: Intonation. Oxford: Oxford University Press.Halliday, Michael A. K. 1970 A Course in Spoken English: Intonation. Oxford: Oxford University Press. observes that in English, a falling intonation contour is associated with certainty and a rising contour with uncertainty. Relating intonation to (procedural) meaning, Clark (1991)Clark, Billy 1991 “Relevance Theory and the Semantics of Non-Declarative Sentences.” PhD dissertation. University College London.Clark, Billy 1991 “Relevance Theory and the Semantics of Non-Declarative Sentences.” PhD dissertation. University College London. proposes that a falling contour encodes the procedural information that the speaker is entertaining the thought described by the proposition expressed (Clark 1991Clark, Billy 1991 “Relevance Theory and the Semantics of Non-Declarative Sentences.” PhD dissertation. University College London.Clark, Billy 1991 “Relevance Theory and the Semantics of Non-Declarative Sentences.” PhD dissertation. University College London., 195), and thus an utterance with a falling contour will tend to be interpreted as an assertion or a directive – in (18a), A is certain that the state of affairs described by the proposition The box goes on the shelf is true, and is either asserting it as a fact or issuing a directive to B to make it so:

Conversely, a rising intonational contour encodes the procedural information that the speaker is not entertaining the thought described by the proposition expressed, so an utterance with a rising contour will tend to be interpreted as a question – in (18b), A indicates that she has not accepted the thought that the box should be placed on the shelf, and would require the addressee to indicate whether it is true.

3.1 Hor ₂₄

Many researchers (Marie 1987Marie, Wilma Vimala 1987 “A Study of Sentence-Final Particles in Singapore English.” Unpublished academic exercise. National University of Singapore.Marie, Wilma Vimala 1987 “A Study of Sentence-Final Particles in Singapore English.” Unpublished academic exercise. National University of Singapore.; Gupta 1992Gupta, Anthea Fraser 1992 “The Pragmatic Particles of Singapore Colloquial English.” Journal of Pragmatics 18: 31–57. Gupta, Anthea Fraser 1992 “The Pragmatic Particles of Singapore Colloquial English.” Journal of Pragmatics 181: 31–57. ; Gan 2000Gan, Lee Meng 2000 “A Study of the ‘Hah’ and ‘Hor’ Particles in Colloquial Singapore English.” Unpublished Honours thesis. National University of Singapore.Gan, Lee Meng 2000 “A Study of the ‘Hah’ and ‘Hor’ Particles in Colloquial Singapore English.” Unpublished Honours thesis. National University of Singapore.; Wee 2004Wee, Lionel 2004 “Reduplication and Discourse Particles.” In Singapore English: A Grammatical Description, ed. by Lisa Lim, 105–126. Amsterdam: John Benjamins. Wee, Lionel 2004 “Reduplication and Discourse Particles.” In Singapore English: A Grammatical Description, ed. by Lisa Lim, 105–126. Amsterdam: John Benjamins. ; Kim and Wee 2009Kim, Chonghyuck, and Lionel Wee 2009 “Resolving the Paradox of Singapore English Hor .” English World-Wide 30: 241–261. Kim, Chonghyuck, and Lionel Wee 2009 “Resolving the Paradox of Singapore English Hor.” English World-Wide 301: 241–261. ) have observed that hor ₂₄ can be used in sentence-final position to solicit agreement for the hor ₂₄-marked utterance, e.g. in (1) (repeated below as [19]) and (20):

In both (19) and (20), the addressee explicitly confirms the speaker’s hor ₂₄-marked utterance, which functions like an interrogative. Almost all of my linguistic consultants (with one exception)88.Nevertheless, the consultant in question finds silence an acceptable response for both (21a) and (21b), so the parallel between a hor-marked utterance and an explicit question still holds true for them as well. agree that, as with explicit interrogatives, while the addressee can disagree with the speaker, he cannot ignore the question altogether:

However, while a hor ₂₄-marked utterance is interpreted like an interrogative, Wee (2004)Wee, Lionel 2004 “Reduplication and Discourse Particles.” In Singapore English: A Grammatical Description, ed. by Lisa Lim, 105–126. Amsterdam: John Benjamins. Wee, Lionel 2004 “Reduplication and Discourse Particles.” In Singapore English: A Grammatical Description, ed. by Lisa Lim, 105–126. Amsterdam: John Benjamins. observes that hor ₂₄ cannot mark actual interrogatives, i.e. polar questions or wh-questions, which is also confirmed by my consultants:

Hor ₂₄ is also generally unacceptable with imperative clauses, unlike hor ₂₁:

As I discuss below in Section 3.3, the utterance in (23) would be interpreted as a question as a result of hor’s rising intonation, which is incompatible with the imperative form and the accompanying context.99.Again, one consultant differs from the rest in finding (23) acceptable as an indirect directive to younger children, interpreting the utterance along the lines of “[You’re supposed to] Stand in line hor ₂₄ ?” (similar to “You’re supposed to stand in line, yes?” in Standard American/British English). The difference in intuition is therefore not in whether the hor ₂₄ -utterance is interpreted as a question, but whether the resultant question can be interpreted as a directive.

Besides marking utterances, hor ₂₄ can also be used as a stand-alone utterance to solicit agreement for a previous utterance, which Gan (2000)Gan, Lee Meng 2000 “A Study of the ‘Hah’ and ‘Hor’ Particles in Colloquial Singapore English.” Unpublished Honours thesis. National University of Singapore.Gan, Lee Meng 2000 “A Study of the ‘Hah’ and ‘Hor’ Particles in Colloquial Singapore English.” Unpublished Honours thesis. National University of Singapore. terms “repair hor”:

In (24), B’s lack of response to A’s statement that their mutual friend is always trying to take advantage of others prompts A to use hor ₂₄ to explicitly solicit a positive response, a function that has also been noted by Wee (2004)Wee, Lionel 2004 “Reduplication and Discourse Particles.” In Singapore English: A Grammatical Description, ed. by Lisa Lim, 105–126. Amsterdam: John Benjamins. Wee, Lionel 2004 “Reduplication and Discourse Particles.” In Singapore English: A Grammatical Description, ed. by Lisa Lim, 105–126. Amsterdam: John Benjamins. .

The use of hor ₂₄ to solicit agreement for an utterance is similar to how tag questions in other varieties of English are used:

Like hor ₂₄, tag questions can also be detached from the preceding utterance:

However, unlike most tag questions,1111.While most tag questions like isn’t it? cannot occur in the middle of a sentence, right is closer to hor ₂₄ in this regard: i. If you all have no problems with Tuesday right then I will reschedule my tuition session. ii. That guy right is very stingy. It would be interesting in future work to explore in greater detail how the semantics of CSE hor ₂₄ and English right may overlap. hor ₂₄ can also occur in the middle of a sentence, e.g. in (27):

Gan (2000Gan, Lee Meng 2000 “A Study of the ‘Hah’ and ‘Hor’ Particles in Colloquial Singapore English.” Unpublished Honours thesis. National University of Singapore.Gan, Lee Meng 2000 “A Study of the ‘Hah’ and ‘Hor’ Particles in Colloquial Singapore English.” Unpublished Honours thesis. National University of Singapore., 37) characterises this as a distinct lexical variant he terms “filler hor”, which occurs at “an intonation break to signal the beginning of new information”. Although Wee (2004)Wee, Lionel 2004 “Reduplication and Discourse Particles.” In Singapore English: A Grammatical Description, ed. by Lisa Lim, 105–126. Amsterdam: John Benjamins. Wee, Lionel 2004 “Reduplication and Discourse Particles.” In Singapore English: A Grammatical Description, ed. by Lisa Lim, 105–126. Amsterdam: John Benjamins. suggests that A is still trying to elicit agreement for the proposition You all have no problems with Tuesday, hor ₂₄ can also occur with noun phrases, as (28) and (29) show:

In (28) and (29), hor ₂₄ is used as a topic marker, which does not require the addressee to agree with the speaker or respond in any way, unlike in (19) and (20). This poses a problem for characterisations of hor where soliciting agreement is an integral part of hor’s semantics, e.g. Wee (2004)Wee, Lionel 2004 “Reduplication and Discourse Particles.” In Singapore English: A Grammatical Description, ed. by Lisa Lim, 105–126. Amsterdam: John Benjamins. Wee, Lionel 2004 “Reduplication and Discourse Particles.” In Singapore English: A Grammatical Description, ed. by Lisa Lim, 105–126. Amsterdam: John Benjamins. , who proposes that hor indicates that the speaker is trying to garner support for the hor-marked proposition, and Ler (2005Ler, Vivien 2005 “An In-Depth Study of Discourse Particles in Singapore English.” PhD dissertation. National University of Singapore.Ler, Vivien 2005 “An In-Depth Study of Discourse Particles in Singapore English.” PhD dissertation. National University of Singapore., 212), who proposes that hor “directs the hearer to focus on information presented and seeks the hearer’s positive response to the proposition”.

In contrast to Wee (2004)Wee, Lionel 2004 “Reduplication and Discourse Particles.” In Singapore English: A Grammatical Description, ed. by Lisa Lim, 105–126. Amsterdam: John Benjamins. Wee, Lionel 2004 “Reduplication and Discourse Particles.” In Singapore English: A Grammatical Description, ed. by Lisa Lim, 105–126. Amsterdam: John Benjamins. and Ler (2005)Ler, Vivien 2005 “An In-Depth Study of Discourse Particles in Singapore English.” PhD dissertation. National University of Singapore.Ler, Vivien 2005 “An In-Depth Study of Discourse Particles in Singapore English.” PhD dissertation. National University of Singapore., Kim and Wee (2009)Kim, Chonghyuck, and Lionel Wee 2009 “Resolving the Paradox of Singapore English Hor .” English World-Wide 30: 241–261. Kim, Chonghyuck, and Lionel Wee 2009 “Resolving the Paradox of Singapore English Hor.” English World-Wide 301: 241–261. characterise hor’s function as a check on whether the addressee shares the speaker’s mental representation, which is specified by the linguistic constituent that hor marks, in the face of an asymmetry in epistemic or deontic authority between the speaker and the addressee. This therefore directly accounts for hor ₂₄’s function of soliciting agreement, as the addressee has epistemic authority over the speaker with respect to the hor-marked proposition in such cases. In addition, Kim and Wee (2009)Kim, Chonghyuck, and Lionel Wee 2009 “Resolving the Paradox of Singapore English Hor .” English World-Wide 30: 241–261. Kim, Chonghyuck, and Lionel Wee 2009 “Resolving the Paradox of Singapore English Hor.” English World-Wide 301: 241–261. analyse hor ₂₄’s function in the middle of utterances as a check “to make sure that the addressee’s mental representation is at least partially similar to that of the speaker’s before continuing with the rest of the utterance” (Kim and Wee 2009Kim, Chonghyuck, and Lionel Wee 2009 “Resolving the Paradox of Singapore English Hor .” English World-Wide 30: 241–261. Kim, Chonghyuck, and Lionel Wee 2009 “Resolving the Paradox of Singapore English Hor.” English World-Wide 301: 241–261. , 258). However, they do not make a distinction between hor ₂₄ and hor ₂₁, and the next section shows how their characterisation of hor fails to account for hor ₂₁’s behaviour with directives.

3.2 Hor ₂₁

Both Marie (1987)Marie, Wilma Vimala 1987 “A Study of Sentence-Final Particles in Singapore English.” Unpublished academic exercise. National University of Singapore.Marie, Wilma Vimala 1987 “A Study of Sentence-Final Particles in Singapore English.” Unpublished academic exercise. National University of Singapore. and Gan (2000)Gan, Lee Meng 2000 “A Study of the ‘Hah’ and ‘Hor’ Particles in Colloquial Singapore English.” Unpublished Honours thesis. National University of Singapore.Gan, Lee Meng 2000 “A Study of the ‘Hah’ and ‘Hor’ Particles in Colloquial Singapore English.” Unpublished Honours thesis. National University of Singapore. characterise hor ₂₁’s function as reducing the harshness of directives (repeated from [3] and [23] respectively):

Specifically, Gan (2000)Gan, Lee Meng 2000 “A Study of the ‘Hah’ and ‘Hor’ Particles in Colloquial Singapore English.” Unpublished Honours thesis. National University of Singapore.Gan, Lee Meng 2000 “A Study of the ‘Hah’ and ‘Hor’ Particles in Colloquial Singapore English.” Unpublished Honours thesis. National University of Singapore. describes hor ₂₁’s function with directives as changing commands or orders into requests.

However, as stated in the introduction, hor ₂₁ generally cannot be used with a directive or request that demands immediate compliance without a pre-existing obligation, such as in (32b):

Looking at the directives that have been documented with hor ₂₁, they can be interpreted as either reminders of pre-existing obligations or requirements, directives that do not require immediate compliance, or negative directives that similarly do not require an overt action to be performed (Examples [33]–[35] respectively):

Kim and Wee (2009)Kim, Chonghyuck, and Lionel Wee 2009 “Resolving the Paradox of Singapore English Hor .” English World-Wide 30: 241–261. Kim, Chonghyuck, and Lionel Wee 2009 “Resolving the Paradox of Singapore English Hor.” English World-Wide 301: 241–261. seem to provide a counter-example to (32) with (36), describing A’s utterance as a demand:

However, I argue that hor ₂₁ is acceptable in (36) because the directive Stand in line is not a command but rather a strong reminder in the discourse context, similar to the directive in (33). Even though (33) and (36) both require their respective addressees to respond immediately by standing in line, they are demands that are only pragmatically inferred given the strong reminder of an already existing requirement to stand in line. In fact, directives that do not require immediate compliance and negative directives that do not require an overt action to be performed can both achieve their pragmatic effect through reminders – a directive that requires an action to be performed at some later time can be interpreted as a reminder for the addressee to perform the required action in the future, while a negative directive can be interpreted as a reminder for the addressee not to perform the action that is being prohibited. In contrast, hor ₂₁ is unacceptable in (32b) as the directive seeks to impose a new obligation on the addressee that would require immediate action, and thus it would be difficult for it to be interpreted as a reminder.1212.A few of my consultants feel that (32b) would be acceptable without a pre-existing obligation only in situations where the speaker was bullying or extorting money from the addressee, in which case hor ₂₁ would likely function as a warning, sharing some similarities with hor ₂₁’s function in (37). While more work will be needed to tease out how hor is used in such situations with highly asymmetrical power dynamics between the interlocutors, I believe that this issue does not significantly affect the analysis of hor in this paper.

The observation that hor ₂₁ is only acceptable with directives that can be interpreted as reminders has not been noted in previous literature, and cannot be accounted for by any of the previous characterisations, which would all predict that hor ₂₁ is acceptable in (32b): Wee (2004)Wee, Lionel 2004 “Reduplication and Discourse Particles.” In Singapore English: A Grammatical Description, ed. by Lisa Lim, 105–126. Amsterdam: John Benjamins. Wee, Lionel 2004 “Reduplication and Discourse Particles.” In Singapore English: A Grammatical Description, ed. by Lisa Lim, 105–126. Amsterdam: John Benjamins. and Ler (2005)Ler, Vivien 2005 “An In-Depth Study of Discourse Particles in Singapore English.” PhD dissertation. National University of Singapore.Ler, Vivien 2005 “An In-Depth Study of Discourse Particles in Singapore English.” PhD dissertation. National University of Singapore. would analyse (32b) straightforwardly as the speaker trying to garner support or seeking the addressee’s positive response to the directive with hor ₂₁, while Kim and Wee (2009)Kim, Chonghyuck, and Lionel Wee 2009 “Resolving the Paradox of Singapore English Hor .” English World-Wide 30: 241–261. Kim, Chonghyuck, and Lionel Wee 2009 “Resolving the Paradox of Singapore English Hor.” English World-Wide 301: 241–261. would analyse (32b) as the speaker trying to use hor ₂₁ to facilitate the successful transfer of the directive from speaker to addressee.

Another problem for previous characterisations of hor is that, instead of marking an imperative directly, a speaker can also use hor ₂₁ to mark a preceding assertion in order to get the addressee to obey the subsequent directive, e.g. in (37):

It is hard to describe hor ₂₁’s function in (37) as garnering support for or seeking the addressee’s positive response to the proposition that the pencil A refers to is her own since this is a clear fact; rather, A seems to be reminding B of the hor ₂₁-marked proposition in order to emphasise the validity of the subsequent imperative, i.e. that B should not touch A’s pencil.

Finally, as with hor ₂₄, all of my consultants agree that hor ₂₁ cannot be used to mark interrogatives, e.g. to remind the addressee of previously asked questions, which one might imagine as a parallel to its function of reminding the addressee of pre-existing obligations:

3.3Intonation

As stated in the introduction, while some researchers have analysed hor as multiple variants distinguished by intonation (Marie 1987Marie, Wilma Vimala 1987 “A Study of Sentence-Final Particles in Singapore English.” Unpublished academic exercise. National University of Singapore.Marie, Wilma Vimala 1987 “A Study of Sentence-Final Particles in Singapore English.” Unpublished academic exercise. National University of Singapore.; Gan 2000Gan, Lee Meng 2000 “A Study of the ‘Hah’ and ‘Hor’ Particles in Colloquial Singapore English.” Unpublished Honours thesis. National University of Singapore.Gan, Lee Meng 2000 “A Study of the ‘Hah’ and ‘Hor’ Particles in Colloquial Singapore English.” Unpublished Honours thesis. National University of Singapore.), others have ignored the different pitch contours hor can be pronounced with, arguing that its different functions are derived entirely from variation in the discourse situation (Wee 2004Wee, Lionel 2004 “Reduplication and Discourse Particles.” In Singapore English: A Grammatical Description, ed. by Lisa Lim, 105–126. Amsterdam: John Benjamins. Wee, Lionel 2004 “Reduplication and Discourse Particles.” In Singapore English: A Grammatical Description, ed. by Lisa Lim, 105–126. Amsterdam: John Benjamins. ; Ler 2005Ler, Vivien 2005 “An In-Depth Study of Discourse Particles in Singapore English.” PhD dissertation. National University of Singapore.Ler, Vivien 2005 “An In-Depth Study of Discourse Particles in Singapore English.” PhD dissertation. National University of Singapore.; Kim and Wee 2009Kim, Chonghyuck, and Lionel Wee 2009 “Resolving the Paradox of Singapore English Hor .” English World-Wide 30: 241–261. Kim, Chonghyuck, and Lionel Wee 2009 “Resolving the Paradox of Singapore English Hor.” English World-Wide 301: 241–261. ). However, examples such as (39) show that intonation does play an important role in the addressee’s interpretation of the particle:

According to my consultants, if hor in (39) has a falling intonation, the utterance can only be interpreted as A reminding B directly that the day is very hot, and so he should remember to drink more water; if it has a rising intonation, the utterance will rather be interpreted as A expressing a shared observation that the day is very hot, which is consistent with hor ₂₄’s function of soliciting agreement on the hor ₂₄-marked utterance that was discussed previously. Nevertheless, if one wishes to preserve a more parsimonious unitary characterisation of hor, there seems at first glance to be a straightforward way to incorporate intonation into Kim and Wee’s (2009)Kim, Chonghyuck, and Lionel Wee 2009 “Resolving the Paradox of Singapore English Hor .” English World-Wide 30: 241–261. Kim, Chonghyuck, and Lionel Wee 2009 “Resolving the Paradox of Singapore English Hor.” English World-Wide 301: 241–261. original characterisation of hor: a rising intonation signals that the addressee has epistemic authority over the speaker and hence hor should be interpreted as a question, while a falling intonation signals that it is the speaker who has epistemic authority and hence hor should be interpreted as a reminder. However, the two interpretations in (39) are tied to their respective intonations even if the epistemic status of the interlocutors is kept identical in both cases, which demonstrates that intonation cannot only be indicating the context of the utterance or the interlocutors’ epistemic status in determining the different functions of hor.

Despite this, I argue in this paper that it is still possible to maintain an analysis of hor that does not involve multiple phonological and semantic variants while still acknowledging intonation as an important factor in how the particle is interpreted. In the following section, I outline an alternative characterisation of hor that maintains its unitary nature while accounting for the effect of intonation on its pragmatic function, as well as other aspects of its use that have been left unexplained in previous analyses.

4.1 Hor ₂₁

As stated in the previous section, hor ₂₁ is only acceptable with directives and assertions that are interpreted as (strong) reminders. In relevance-theoretic terms, for a proposition to be remembered, it must be kept accessible. As such, I propose that hor ₂₁ functions as an instruction for the addressee to keep the hor ₂₁-marked proposition accessible in his cognitive environment.

To illustrate, consider (41):

Although Ler (2005)Ler, Vivien 2005 “An In-Depth Study of Discourse Particles in Singapore English.” PhD dissertation. National University of Singapore.Ler, Vivien 2005 “An In-Depth Study of Discourse Particles in Singapore English.” PhD dissertation. National University of Singapore. analyses (41) as a directive, i.e. A is requesting B to help her get a copy of Today, the difference between (42a) and (42b) shows that this cannot be the case:

As stated in Section 2.2, imperatives are analysed as encoding descriptions of situations that are potential and desirable, i.e. (43a):

Whether B obeys A’s directive will depend on how this description interacts with the other premises that B has in his cognitive environment, as outlined in (43b). Since B would try to do what A wants in order to make her happy because A is B’s wife, given Premises 1 and 2, B would arrive at the conclusion that he should obey A and get her the paper. I propose that hor ₂₁ acts on the encoded proposition p in (43a), functioning as a reminder by instructing the addressee to keep p accessible. The directive “Get Today hor ₂₁” is therefore interpreted not as a direct instruction for B to get the paper Today, as in (43), but merely for him to remember that A wants him to get the paper for her. Since the latter only requires a verbal response to indicate that the reminder has been received, rather than an action to be performed immediately, (44b) is thus unacceptable as a response:

This downgrading of the illocutionary force of hor ₂₁-marked utterances from directives to reminders also explains Marie’s (1987)Marie, Wilma Vimala 1987 “A Study of Sentence-Final Particles in Singapore English.” Unpublished academic exercise. National University of Singapore.Marie, Wilma Vimala 1987 “A Study of Sentence-Final Particles in Singapore English.” Unpublished academic exercise. National University of Singapore. and Gan’s (2000)Gan, Lee Meng 2000 “A Study of the ‘Hah’ and ‘Hor’ Particles in Colloquial Singapore English.” Unpublished Honours thesis. National University of Singapore.Gan, Lee Meng 2000 “A Study of the ‘Hah’ and ‘Hor’ Particles in Colloquial Singapore English.” Unpublished Honours thesis. National University of Singapore. characterisation of hor ₂₁ as a politeness device that softens commands, which is supported by the intuitions of most of my consultants. Instead of having to obey the directive because of A’s authority over B, reminding B of the action’s desirability to A shifts B’s motivation for bringing it about to A’s relationship with B. However, similar to how the hor ₂₁-marked reminder “Stand in line hor ₂₁” is still interpreted as a demand in (36) above, some of my consultants commented that this also depends on the discourse context, as the hor ₂₁-marked reminder “Get Today hor ₂₁” may be interpreted as a warning if B had forgotten the obligation several times before, or if A believes that B may forget it later, which would come across as being more forceful and less polite instead.

The characterisation of hor ₂₁ as an instruction for the addressee to keep the hor ₂₁-marked proposition accessible in his cognitive environment also accounts for how it can be used to justify or emphasise a directive even when it is attached to a preceding declarative utterance, e.g. in (37), repeated below as (45):

Following the analysis above, we see that the reminder the hor ₂₁-marked declarative becomes in (45) serves as a premise that justifies the subsequent directive – since this is A’s pencil, B should not touch it. This is also similar to the hor ₂₁ -marked version of (39), repeated below as (46):

As with (45), A indicates with (46) that B should remember to drink more water because the day is very hot.

4.2 Hor ₂₄

Like hor ₂₁, I propose that when hor ₂₄ is used in sentence-final position, the addressee interprets hor first as indicating that the proposition it marks is accessible. However, the rising intonation then turns it into a question asking the addressee to acknowledge whether the hor-marked proposition is accessible in his cognitive environment.

To illustrate, consider (1) (repeated below as [47]):

In (47), A uses hor ₂₄ to check if B agrees with her on the hor ₂₄-marked utterance. Without taking intonation into account, hor marks a proposition as being accessible; the rising intonation then indicates that the thought of the given proposition p being accessible (to B) is desirable or relevant to the speaker if true. However, since the speaker A does not know whether the proposition that there were no shops on the street previously is accessible to B unless B tells her so, B would have to acknowledge whether the proposition is accessible in his cognitive environment, or if it is not accessible because it is in fact false.

Following Sperber and Wilson’s (1995)Sperber, Dan, and Deirdre Wilson 1995 Relevance: Communication and Cognition. Oxford: Blackwell.Sperber, Dan, and Deirdre Wilson 1995 Relevance: Communication and Cognition. Oxford: Blackwell. description of a person’s cognitive environment as consisting of “not only all the facts that he is aware of, but also all the facts that he is capable of becoming aware of, in his physical environment” (Sperber and Wilson 1995Sperber, Dan, and Deirdre Wilson 1995 Relevance: Communication and Cognition. Oxford: Blackwell.Sperber, Dan, and Deirdre Wilson 1995 Relevance: Communication and Cognition. Oxford: Blackwell., 39), propositions in a person’s cognitive environment are generally assumed in a neutral context to be true. By checking the accessibility of a proposition in the addressee’s cognitive environment using hor ₂₄, rather than asking whether it is true, the speaker indicates that she assumes that the proposition is true; since checking the accessibility of a true proposition in one’s cognitive environment would itself make it easier to recall and hence more accessible, I argue that the speaker would therefore expect a positive response to her hor ₂₄-marked utterance, which accounts for its bias in soliciting agreement.

In support of this, a search in the Same Room transcripts of the conversational section of the NSC, where participants talk face-to-face, shows that, while the addressee can still disagree with the hor ₂₄-marked utterance, the speaker is usually validated in her expectation of a positive response. Exchanges where hor occurs at the end of the line in the transcription (to exclude instances of hor as a topic marker), followed by a response by the other interlocutor, were extracted from the Same Room conversations using a script in the R programming platform (R Core Team 2021R Core Team 2021 R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. https://​www​.R​-project​.org/R Core Team 2021 R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. https://​www​.R​-project​.org/). They were then manually inspected to remove instances involving ya hor, an expression indicating agreement that does not seem to be fully compositional (the alternative yes hor rarely appears in the corpus), and orh hor, an expression of accusation that again seems to be a calqued phrase. The corresponding audio recordings were also consulted to determine where hor ₂₄ is used. Out of the forty-eight examples of sentence-final hor ₂₄ identified in the Same Room conversations,1313.In these forty-eight examples, hor with a rising pitch contour is the only particle in sentence-final position. However, an additional twenty-one examples have the particle with a rising pitch contour occurring after another particle like lah. While an analysis of how hor functions in these particle combinations is beyond the scope of this paper (but see Boo et al. 2023Boo, Ashley, Lee Junwen, and Tan Ying-Ying 2023 “Particle Stacking in Singlish: New Data from the National Speech Corpus.” Lingua 287: 103513. Boo, Ashley, Lee Junwen, and Tan Ying-Ying 2023 “Particle Stacking in Singlish: New Data from the National Speech Corpus.” Lingua 2871: 103513. for a discussion of particle combinations in CSE), the other interlocutor explicitly agrees with the speaker in thirteen of those examples, compared to one case of explicit disagreement, which is consistent with the examples where hor is the only sentence-final particle in the utterance. twenty-six were followed by explicit agreement by the other interlocutor, e.g. Mm or Ya/Yeah/Yes, etc., while only six were followed by explicit disagreement.1414.Besides showing explicit agreement or disagreement, in eleven other examples, the other interlocutor implicitly agrees with the speaker by building on their utterance, e.g.: (i)(Context: A and B are talking about a sushi franchise in Singapore that B likes.) A:right not many in Singapore hor ₂₄ outlets hor ₂₄ B:tried the one I tried is my uh City Hall lah (NSC 3181) While B does not explicitly agree with A’s hor ₂₄-marked utterance that there are not many outlets of the sushi franchise in Singapore, B does not contradict A either, but rather builds on what she says to clarify that the specific outlet she had been to was at City Hall. Finally, the other interlocutor asked the speaker in two other examples to repeat or clarify what they said, while in the remaining three examples, it was unclear which of the speaker’s previous utterances the other interlocutor was responding to.

Finally, to account for sentence-medial hor ₂₄’s function as a topic marker, I propose that the particle also marks a desirable thought. However, it is desirable to the addressee rather than the speaker, and instead of deriving relevance from the thought that p is accessible, the addressee derives relevance from the checking of p’s accessibility itself. In this way, it functions in the same way as the use of the question “Remember Tom?” in (48):

In (48), the question conveys to the addressee that the thought that he remembers Tom is desirable either to the speaker or himself. However, the fact that the speaker does not wait for the addressee’s answer but moves on directly to her next point suggests that the thought is desirable not to herself but to the addressee. By asking the addressee to ascertain the truth value of the question, the speaker intends the memory of Tom (together with other associated memories) to become accessible to him. This thus allows the addressee to process the next utterance “Well, he’s just bought a motorbike” in the context of his existing knowledge of Tom, thereby reducing the processing costs required for the addressee to a minimum and achieving optimal relevance.

I propose that the use of hor ₂₄ in (29) (repeated below as [49]) can be analysed in a similar way:

In ascertaining whether the proposition There is that guy in (49) is accessible, the addressee causes assumptions about the man to become accessible, providing the contextual information required to process the subsequent utterance “very stingy” – the corresponding utterance in Standard American/British English, similar to (48), would be something like “Remember / You know that guy? He is very stingy”.

The analysis of sentence-medial hor ₂₄ in this paper is similar to Kim and Wee’s (2009)Kim, Chonghyuck, and Lionel Wee 2009 “Resolving the Paradox of Singapore English Hor .” English World-Wide 30: 241–261. Kim, Chonghyuck, and Lionel Wee 2009 “Resolving the Paradox of Singapore English Hor.” English World-Wide 301: 241–261. analysis of “filler hor” as a checking mechanism for the speaker “to make sure that the addressee’s mental representation is at least partially similar to that of the speaker’s before continuing with the rest of the utterance” (Kim and Wee 2009Kim, Chonghyuck, and Lionel Wee 2009 “Resolving the Paradox of Singapore English Hor .” English World-Wide 30: 241–261. Kim, Chonghyuck, and Lionel Wee 2009 “Resolving the Paradox of Singapore English Hor.” English World-Wide 301: 241–261. , 258). However, since sentence-medial hor ₂₄ does not require a response by the addressee, as stated in Section 3.1, it is not clear how the addressee would be able to ascertain the speaker’s mental representation through the hor ₂₄-marked utterance alone, in order to assess if his mental representation is similar to hers or not. In contrast, checking if a specific proposition marked by hor ₂₄ is accessible is a much more constrained operation, which I argue recommends the characterisation of hor in this paper over Kim and Wee’s (2009)Kim, Chonghyuck, and Lionel Wee 2009 “Resolving the Paradox of Singapore English Hor .” English World-Wide 30: 241–261. Kim, Chonghyuck, and Lionel Wee 2009 “Resolving the Paradox of Singapore English Hor.” English World-Wide 301: 241–261. . In addition, as stated in Section 3.1, Kim and Wee’s (2009)Kim, Chonghyuck, and Lionel Wee 2009 “Resolving the Paradox of Singapore English Hor .” English World-Wide 30: 241–261. Kim, Chonghyuck, and Lionel Wee 2009 “Resolving the Paradox of Singapore English Hor.” English World-Wide 301: 241–261. characterisation of hor does not differentiate between the different intonations that the particle can be used with, and so does not correctly account for how hor ₂₁ is used with directives.