Default Semantics Research Articles

Safe-by-default Concurrency for Modern Programming Languages

Modern “safe” programming languages follow a design principle that we call safety by default and performance by choice . By default, these languages enforce important programming abstractions, such as memory and type safety, but they also provide mechanisms that allow expert programmers to explicitly trade some safety guarantees for increased performance. However, these same languages have adopted the inverse design principle in their support for multithreading. By default, multithreaded programs violate important abstractions, such as program order and atomic access to individual memory locations to admit compiler and hardware optimizations that would otherwise need to be restricted. Not only does this approach conflict with the design philosophy of safe languages, but very little is known about the practical performance cost of providing a stronger default semantics. In this article, we propose a safe-by-default and performance-by-choice multithreading semantics for safe languages, which we call volatile -by-default . Under this semantics, programs have sequential consistency (SC) by default, which is the natural “interleaving” semantics of threads. However, the volatile -by-default design also includes annotations that allow expert programmers to avoid the associated overheads in performance-critical code. We describe the design, implementation, optimization, and evaluation of the volatile -by-default semantics for two different safe languages: Java and Julia. First, we present V BD-HotSpot and V BDA-HotSpot, modifications of Oracle’s HotSpot JVM that enforce the volatile -by-default semantics on Intel x86-64 hardware and ARM-v8 hardware. Second, we present S C-Julia, a modification to the just-in-time compiler within the standard Julia implementation that provides best-effort enforcement of the volatile -by-default semantics on x86-64 hardware for the purpose of performance evaluation. We also detail two different implementation techniques: a baseline approach that simply reuses existing mechanisms in the compilers for handling atomic accesses, and a speculative approach that avoids the overhead of enforcing the volatile -by-default semantics until there is the possibility of an SC violation. Our results show that the cost of enforcing SC is significant but arguably still acceptable for some use cases today. Further, we demonstrate that compiler optimizations as well as programmer annotations can reduce the overhead considerably.

Should Special Science Laws Be Written into the Semantics of Counterfactuals?

Abstract Adam Elga has presented an anti-thermodynamic process as a counterexample to Lewis’s default semantics for counterfactuals. The outstanding reaction of Jonathan Schaffer and Boris Kment is revisionary. It sacrifices Lewis’s aim of defining causation in terms of counterfactual dependence. Lewis himself suggested an alternative: «counter-entropic funnybusiness» should make for dissimilarity. But how is this alternative to be spelled out? I discuss a recent proposal: include special science laws, among them the laws of thermodynamics. Although the proposal fails, it serves to uncover the limits of Elga’s example.

When Conditionals Lack Conditionality: The Semantics of Biscuit Conditionals in some Selected Texts from the Holy Quran and the Qudsi Hadiths: A Default Semantics-Theoretic Perspective

Drawing on Jaszczolt’s (2005) Default Semantics theory, and inspired by Austin’s (1961, p. 158) original example: “There are biscuits on the sideboard if you want some,” this paper attempts a semantic analysis of Biscuit Conditionals (BCs) in some selected texts from the Holy Quran and the Qudsi Hadiths, with the intention to explore how this type of conditionals communicates discourse meanings other than conditionality. The main objective of this paper is to show how BCs can semantically be represented to communicate non-conditional meanings within the framework of Default Semantics, a contextualism-oriented and truth-conditional approach that aims to identify the primary meanings pertaining to utterances. Four research questions are presented here: first, what are the different types of BCs presented in the selected texts from the Holy Quran and the Qudsi Hadith? Second, why are the selected constructions considered Biscuit Conditionals not any other type of conditionals? Third, what are the didactic purposes conveyed by BCs in the selected data? Fourth, how are BCs semantically represented to convey non-conditional meanings within the framework of a religion-oriented type of communication, and in light of a theory of discourse meaning? In light of Default Semantics, the paper concludes that Biscuit Conditionals in the selected data lack conditionality on the level of their intended meanings, and assign the discourse functions of fact-confirmers, information-notifiers, sympathy-instigators, invitation-motivators, discourse-relievers and offer-activators that function as conduits of religious didacticism

Representing the cognitive construal of Chinese first-person singular reference in discourse

This study aims to explore the Chinese first-person singular reference in discourse, and exhibit the complexity of its usage. Default Semantics is applied in the construal of Chinese first-person singular reference phenomenon taken as an act of communication. After analyzing the examples elicited from the Lancaster Chinese Corpus, we find that the discourse meanings of Chinese first-person singular pronoun wǒ are not limited to self-reference, a variety of additional or non-deictic meanings are conveyed by it in different contexts, and many non-deictic expressions can also be employed to convey self-referential meanings. The construal of Chinese first-person singular reference discourse meaning results from the interaction of pragmatics with grammar and semantics. Thus, the interface between semantics and pragmatics should be rethought with respect to specific language interpretation.

A Lexical Semantics for Refugee, Asylum Seeker and Boat People in Australian English

The terms refugee, asylum seeker and boat people are of particular prominence in the Australian discourse surrounding immigration policy, and are widely used in day-to-day conversation among Australians. Despite their frequency of use, a lexico-semantic study of the terms has not been carried out to date. This paper fills this gap by proposing a semantic analysis of them. The study is based on a corpus created from online comments to the Australian television programme Go Back To Where You Came From (Season 1, SBS 2011). After introducing the data and analytical framework—object-oriented semantics—we discuss the terms’ lexical semantics. While the discussion of immigration issues is emotionally laden, our results suggest that the default semantics of the terms do not include evaluative components. Rather, speakers tend to evaluate the agreed-upon semantic specifications differently depending on their political views. We show how each term represents a specific node in a network of concepts for translocating individuals, but may in context also be applied to neighbouring nodes that lack a lexicalization. While the terms are seemingly used interchangeably, our analysis instead emphasizes the influence of the underlying conceptual structure and the resulting constrained plasticity of nominal meaning in context.

Précis by Capone in response to Zhang and Zhang

In this paper I respond to some provocative remarks by Zhang and Zhang (2017) and I argue that their claim that there is no compatibility between Default semantics and Relevance theory is exaggerated. In Capone (2011a,b), I provided a synthetic approach which did not presuppose that we necessarily slavishly adhere to all details of Jaszczolt's or Relevance Theory. In that paper, I tried to make sense of the idea of modularity of mind. However, those ideas could also be integrated into a theory of neural networks. In this paper, I point out, however briefly, that it makes sense to connect the generalizations offered by Default Semantics with the role played by Broca’ s area in cognition (here I follow Pennisi and Falzone's (2016) ideas) (see also Capone, 2017).

Time, perspective and semantic representation

abstractI discuss the perspectival nature of temporality in discourse and argue that the human concept of time can no more be dissociated from the perspectival thought than the concept of the self can. The corollary of this observation is that perspectival temporality can no more be excluded from the semantic representation than the notion of the self can: neither can be reduced to the bare referent for the purpose of semantic representation if the latter is to retain cognitive plausibility. I present such a semanticquaconceptual approach to temporal reference developed within my theory of Default Semantics. I build upon my theory of time as epistemic modality according to which, on the level of conceptualquasemantic building blocks, temporality reduces to degrees of detachment from the certainty of thehereand thenow. I also address the questions of temporal asymmetry between the past and the future, and the relation between metaphysical time (timeM), psychological time (timeE, where ‘E’ marks the domain of epistemological enquiry), and time in natural language (timeL), concluding that the perspective-infused timeEand timeLare compatible with timeMof mathematical models of spacetime: all are definable through possibility and perspectivity.

Explicature versus default meaning: A response to Alessandro Capone's Default Semantics and the architecture of the mind

Alessandro Capone (2011) makes an attempt to provide a unified picture for Relevance Theory and Default Semantics by supporting the idea of the division of labor between these two approaches; that is, both are needed and are complementary subparts of a superordinate theory. We argue against Capone's account by recapitulating the two concepts of explicatures and default meanings, and maintain that Relevance Theory and Default Semantics are both derived from reconceptualizing Grice's generalized conversational implicature. These theories provide different approaches to such additional meanings of uttered sentences by holding different theoretical positions. Therefore, it is hard to draw a picture of them as unified.

Towards a pragmatic category of conditionals

In this paper, we present the benefits of regarding conditionality as a pragmatic phenomenon as compared with approaches based on the syntactic category of a conditional sentence. We propose a pragmatic category of conditionality and justify it using theoretical arguments supported with examples from our database collected from the International Corpus of English-GB. Next, we demonstrate how conditional utterances that pertain to a variety of syntactic constructions can be represented in Default Semantics, a contextualist, truth-conditional approach to utterance meaning. We identify six types of such constructions, using the dimensions of (i) primary vs. secondary meaning (PM/SM index) and (ii) meaning conveyed through sentence structure vs. meaning conveyed at the level of merger representation (WS/Σ index). It is concluded that in view of the diversity of constructions through which conditional thoughts are expressed, conditionality is best regarded as a pragmatic (and as such conceptual) category. Finally, we comment on the status of this claim as a potential semantic/pragmatic universal.

Execution levels for aspect-oriented programming: Design, semantics, implementations and applications

In aspect-oriented programming (AOP) languages, advice evaluation is usually considered as part of the base program evaluation. This is also the case for certain pointcuts, such as if pointcuts in AspectJ, or simply all pointcuts in higher-order aspect languages like AspectScheme. While viewing aspects as part of base level computation clearly distinguishes AOP from reflection, it also comes at a price: because aspects observe base level computation, evaluating pointcuts and advice at the base level can trigger infinite regression. To avoid these pitfalls, aspect languages propose ad-hoc mechanisms, which increase the complexity for programmers while being insufficient in many cases. After shedding light on the many facets of the issue, this paper proposes to clarify the situation by introducing levels of execution in the programming language, thereby allowing aspects to observe and run at specific, possibly different, levels. We adopt a defensive default that avoids infinite regression, and gives advanced programmers the means to override this default using level-shifting operators. We then study execution levels both in practice and in theory. First, we study the relevance of the issues addressed by execution levels in existing aspect-oriented programs. We then formalize the semantics of execution levels and prove that the default semantics is indeed free of a certain form of infinite regression, which we call aspect loops. Finally, we report on existing implementations of execution levels for aspect-oriented extensions of Scheme, JavaScript and Java, discussing their implementation techniques and current applications.

‘The Best Architect Designed This Church’: Definite Descriptions in Default Semantics

Jaszczolt examines the default semantic analysis of ‘the best architect designed this church’, uttered when standing in front of El Temple de la Sagrada Família in Barcelona. At first sight, Jaszczolt's conclusion that the cognitive default reading of the sentence is ‘Antoni Gaudí designed El Temple de la Sagrada Família, and the speaker believes him to be the best architect’ looks right; but is it in fact an appropriate semantic analysis, even given Jaszczolt's pragmatics-rich approach? Our knowledge of what the particular speaker is likely to know is our only guide here. We will make different assumptions for the speaker who is a four-year-old child, George W. Bush, or a native of Barcelona (and so on). The same goes for the addressee: an adult addressing a four-year-old child will very likely follow up by identifying the architect. We only know whether speaker and hearer correctly identify the architect from the co-text, and this requires considerably more inferencing from contextual and encyclopaedic data than Jaszczolt allows. This paper criticizes some of the assumptions of Default Semantics and suggests some emendations to the theory, including additional machinery showing a mapping from the words uttered to the intended meaning in the case of ‘the best architect designed this church’.

The Attributive/Referential Distinction, Pragmatics, Modularity of Mind and Modularization

In this paper I deal with the attributive/referential distinction. After reviewing the literature on the issue, I adopt Jaszczolt's view based on default semantics. I relate her view to Sperber and Wilson's Principle of Relevance. I argue in favour of the modularity hypothesis in connection with pragmatic interpretations. I also discuss the issue of modularization à la Karmiloff-Smith in connection with default inferences and, in particular, referential readings of NPs. I reply to some considerations by Cummings and use data from referential/attributive uses of NPs to show that the modularity hypothesis is defensible.

Default Semantics and the architecture of the mind

In this paper, I explore the relationship between Relevance Theory and Jaszczolt's Default Semantics, framing this debate within the picture of massive modularity tempered by the idea of brain plasticity ( Perkins, 2007). While Relevance Theory focuses on processing (see cognitive efforts and contextual effects interplay), Default Semantics focuses on types of sources from which addressees draw information and types of processes that interact in providing it. In particular, I argue that Relevance Theory interacts with default semantics by standardizing inferences which are ultimately compressed (to use a term by Bach, 1998) into a default semantics. I briefly discuss potential obstacles to the idea of default semantics coming from the experimental pragmatics literature (e.g. Noveck and Sperber, 2007; Breheny et al., 2005) and I support further the idea of the division of labor default inferences and the inferences derivable through the Principle of Relevance. In the end, I compare Relevance Theory and Default Semantics, in an attempt to come to a more unified picture.

Temps comme degrés d'engagement épistémique

This article addresses the question as to whether time is a primitive concept or rather is composed of conceptually more basic building blocks. After a brief analysis of tense-time mismatches with examples from English, Polish, Thai and Swahili, I present a hypothesis that time is conceptualized in terms of degrees of epistemic modality. Expressions with future, present and past reference are ordered on scales of epistemic commitment. I demonstrate that the theory of Default Semantics has no difficulty with representing tense-time mismatches in that it reflects the fact that information about temporality is conveyed via a variety of processes, some of them pertaining not to the processing of the lexicon or grammar but even to pragmatic inference. The theory also gives support to the thesis of time as modal detachment.

Referring as a pragmatic act

Descriptions of pragmemes and pragmatic acts in Mey (2001) and Capone (2005) seem to reduce pragmemes to little more than illocutionary or perlocutionary types. Emphasizing the importance of context to the proper delivery and interpreting of speech acts scarcely justifies introducing new terminology: the importance of common ground (discourse context, situation of utterance, and input from relevant encyclopaedic knowledge) in determining the forms and meanings of speech acts has been recognized, if not by Austin (1962), at least since Bach and Harnish (1979). My paper expands the notion of ‘pragmatic act’ to include other things done with language than performing illocutionary acts, even if the acts I shall identify can rightly be included as components of speech acts. I examine some assumptions of Default Semantics ( Jaszczolt, 2005) as a means of eliciting what certain pragmemes may be expected to accomplish. In particular I shall consider (1) whether intentionality counts as a pragmatic act—as perhaps an instance of contextualized adaptive behaviour; (2) whether a speaker's referring counts as a pragmatic act; and (3) whether the hearer's understanding of an utterance, being the social default in a given situation, counts as a pragmatic act. And, for those that are pragmatic acts, I suggest a programmatic description for the pragmeme that will define such acts. I shall also propose some emendations to the theory of Default Semantics.

Situated temporal reference: A case for compositional pragmatics?

The paper proposes an analysis of temporal reference in discourse which assumes situated acts of communication and their abstract equivalents ( pragmemes, Mey, 2001, 2007) and suggest a formal compositional account of acts of communication in the framework of Default Semantics (DS, Jaszczolt, 2005, 2009, 2010). DS models abstract, situated acts of communication but in addition offers semantic, truth-conditional representations of situated meanings, in that, in accordance with the principles of contextualism, truth-conditional representations allow for the contribution of the output of pragmatic processing. This approach is then exemplified in the representation of temporality.

Default semantics. Foundations of a compositional theory of acts of communication: K. Jaszczolt, OUP, Oxford, 2005, 277 pp.

Default semantics. Foundations of a compositional theory of acts of communication: K. Jaszczolt, OUP, Oxford, 2005, 277 pp.

Variadic function and pragmatics-rich representations of belief reports

Truth-conditional pragmatics (TCP, Recanati, 2002, 2003, 2004) advocates so-called ‘top-down’ pragmatic processes that contribute to the truth-conditionally evaluable representation of meaning while not being grammatically controlled. In this paper I propose an analysis of meaning of utterances expressing propositional attitude reports, starting with the assumptions of TCP and in particular employing the device of a variadic function ( Recanati, 2002, 2005a). Propositional attitude reports are notorious for the problem with accounting for their meaning in that, being a sub-species of intensional contexts, they are sensitive to the substitution of coreferential expressions. Such a substitution may change the meaning of the construction. Following my analysis proposed in Jaszczolt (2005a), I employ Recanati's concept of variadic adicity to the description of the attitudinal predicate. In particular, I look at expressions of the form ‘A believes that B φs’ and suggest that in order to account for the contribution of the description of B to the meaning of the report, one has to postulate an argument slot for the belief predicate that is filled in by the relevant aspects of the mode of presentation of B when appropriate, and left unfilled in other cases. In effect, this amounts to the variable adicity of the belief operator, varying between two and three arguments: the holder of the belief, the proposition, and on some occasions the mode of presentation under which the proposition is believed. I also comment on the compatibility of the hidden-indexical theory of belief reports with the device of variadic function: by putting the two together we obtain a much more successful account of these problematic constructions. Next, I address the pertinent issue of the division of labour between semantics and pragmatics, discussing it in the example of the possibility of a formal account within the overall confines of TCP. I make use of Recanati's pragmatics-rich notion of compositionality ( Recanati, 2004:132) and suggest that compositionality is to be sought on the level of representations of utterance meaning that combines information about meaning coming from different sources, such as the lexicon, sentence structure, subdoxastic enrichment, and conscious pragmatic inference. In the following section, I demonstrate how these insights can be put together in the framework of Default Semantics ( Jaszczolt, 2005a) that rests on this assumption of such higher-level compositionality. I conclude with a brief discussion of the impact of this analysis on the minimalism/contextualism/meaning eliminativism debate in post-Gricean pragmatics.

Meaning merger: Pragmatic inference, defaults, and compositionality

Kecskes: Linguistic underspecification of utterance content is widely accepted across different frameworks, including the neo-Gricean approaches (cf. Horn 2005; Levinson 2000) and relevance theory (Carston 2002, 2005; Sperber & Wilson 1986, 1995). There is also an agreement that if linguistic underdeterminacy is given, pragmatic inference is required if a hearer is to recover a speaker's meaning successfully. In your Default Semantics, you reject the idea of underdetermined semantic representation, and offer an alternative approach in which semantic representation is established with the help of intentions in communication. This means that intentions “intrude” into the semantic representation, and the semantic and pragmatic components are interwoven. What are the advantages of this one-level semantics as opposed to the modular view? It can be argued that, in a way, Default Semantics is also a modular approach because intentions can be considered preverbal thoughts generated in the “conceptualizer” and linguistically shaped in the “formulator” using Levelt's terminology (Levelt 1989, 1999). Do you agree with this line of thinking?

Against ambiguity and underspecification: evidence from presupposition as anaphora

I propose a critical assessment of both the underspecification (sense-generality) and ambiguity positions. I argue that within the frameworks that espouse the dynamic approach to meaning construction, a non-ambiguous interpretation ensues. My Default Semantics that merges the logical form with the output of the intentionality of the corresponding mental states [Default semantics, pragmatics and intentions in Turner, ed. (1999); Disourse, Beliefs and Intentions (1999); The default-based context-dependence of belief reports in Jaszczolt, ed. (2000)] demonstrates that the concept of underspecification is redundant in the case of definite descriptions. In the case of apparently semantically ambiguous referring expressions, the speaker's intentions interact with the logical form of the sentence and produce a unique propositional representation. The problem of ambiguity does not arise because intentions ‘intrude’ into the semantic representation. The view is compatible with pragmatic intrusionism of dynamic-semantic approaches such as Discourse Representation Theory. In this paper I present an argument against underspecification that comes from the approach to presupposition as anaphora by van der Sandt [J. Semantics (1992)] as supplemented by Krahmer and van Deemter [J. Semantics (1998)]. Van der Sandt suggests that presuppositional expressions are anaphoric expressions. When accounted for in terms of a dynamic semantic theory, they exhibit the capacity to bind to an antecedent or to accommodate in the relevant context, and these processes are aided by pragmatic factors. I propose an amendment to van der Sandt's and Krahmer and van Deemter's approach by demonstrating that where multiple representations are viable, they are ordered on the scale of preference that does not allow for ambiguities. I replace van der Sandt's admittance of some degree of ambiguity of presuppositional anaphors with a scale of salience of possible anaphors, ranging from the default to the most unlikely one. Underspecification is not denied its theoretical status: if the theoretical discussion focuses on the theoretical question as to to what extent sentences have to be disambiguated for logical reasoning to proceed [van Deemter, J. Semantics (1998)], distinguishing the stage of underspecification is justified. Instead, I question the need for distinguishing underspecified representation as a separate stage in utterance interpretation, at least in the fragments of English that I have tested so far.