Over the last few decades, much work in formal pragmatics has focused on the issue of whether the syntactic complexity of an expression restricts the range of meanings an expression can have and the amount of semantic variability it permits. In this course, we will critically examine some recent debates around this issue, focusing especially on the idea that the alternatives used to strengthen the meaning of an expression cannot exceed it in complexity (Katzir 2007, Fox & Katzir 2011).
Effects of structure on strengthening are particularly well-studied in the domain of scalar inferences: (1) can be strengthened on the basis of an equally complex `all’-alternative as in (1-a), but not on the basis of a more complex negative alternative (1-b). (2) illustrates a similar asymmetry in the domain of Maximize Presupposition (Percus 2006, Sauerland 2008a a.o.): The oddness of (2-a) is presumably due to the existence of an alternative with a stronger presupposition that superficially seems equally complex; (2-b) shows that more complex alternatives with stronger presuppositions do not trigger oddness.
(1) Ann ate some of the leftovers.
a. equally complex alternative: Ann ate all of the leftovers
-> available strengthened reading: ‘Ann ate some, but not all of the leftovers’
b. more complex alternative: Ann did not eat all of the leftovers
-> unavailable strengthened reading: ‘Ann ate all of the leftovers’
(2) a. # John broke all of his arms (alternative with stronger presupposition: John broke both of his arms)
b. John broke all of his fingers (alternative with stronger presupposition: John broke all ten of his fingers)
Examples like (1) and (2) are within the scope of the traditional approach that alternatives must be obtained by replacing scalar items with their scalemates (Horn 1972, 1989, Gazdar 1979). The idea that the structural complexity of an expression more generally puts an upper bound on its alternatives (Katzir 2007, Fox & Katzir 2011; henceforth the “upper bound approach”) removes the need for stipulated scales and also permits alternatives derived by deletion, but its generality and broader scope also makes it susceptible to new challenges. A closer look at these challenges might teach us something about which aspects of syntactic structure-building are `visible’ to the semantics/pragmatics interface. Conversely, if we develop a better understanding of syntactic constraints on alternatives, we might eventually turn strengthening phenomena into a diagnostic tool for syntactic structure. Therefore, the topic should be of interest also to students whose main research interests are not in pragmatics.
After introducing the upper bound approach and some of its applications, I will try to explore the following unresolved issues, in the hope of stimulating future work by students on these phenomena:
1) Recent work has identified challenges for the upper bound approach that involve both apparently available alternatives of higher complexity and apparently unavailable alternatives of sufficiently low complexity. Should we respond to these challenges by rejecting structural conditions on alternatives altogether (Buccola et al. 2022, Schwarz & Wagner 2024a,b) or should we rethink the nature of the constraints on possible alternative sets, while still taking them to be structure-sensitive (Haslinger & Schmitt 2025, Bar-Lev et al. 2025)?
2) Assuming that there is some role for structure in constraining alternatives, are there syntactic complexity asymmetries that are too `minor’ for the pragmatics to see? Particularly interesting here are Maximize Presupposition effects triggered by phi-features (e.g. Sauerland 2008b)—for instance, local person pronouns should count as alternatives of third person pronouns, even though a decompositional approach to the morphosyntax of person would predict the former to be more complex. Analogous cases arguably arise in the scalar implicature domain. Can we characterize these cases syntactically? Do lexicalization patterns (e.g. shared vs. separate exponents for two heads) matter for the purposes of determining alternatives? Could notions of locality matter?
3) The upper bound approach differs from a theory based on lexicalized scales in that it extends to strengthening inferences based on alternatives to open-class items, where world knowledge and conceptual hierarchies partly determine sets of alternatives at the ‘same specificity level’ , as in (3). We will see that this kind of strengthening shows parallels to scalar inferences that support a unified approach (see e.g. Meyer 2013), but also does not appear to be subject to clear constraints on the internal complexity of the open-class expressions. This raises the suspicion that, more generally, certain subexpressions are ‘opaque’ for the purpose of determining the relative complexity of alternatives, and the question of whether we can predict when this `opacity’ occurs.
(3) Ann (only) went to France.
strengthening inference: Ann didn’t go to Germany, didn’t go to Croatia, etc.
Time permitting, I will also discuss an application of syntactic complexity in pragmatics that is not directly about strengthening: the idea that there is a defeasible constraint on expressions that are truth-conditionally redundant (i.e. have a simpler alternative with contextually equivalent truth conditions), and violations of this constraint are acceptable only if they help us avoid a violation of another (Quantity- or Manner-based) pragmatic maxim. This idea has been applied e.g. by Solt (2018) to the distribution of approximators (about 20 as opposed to 20) and by Haslinger (2024) to correlations between complexity and imprecision in plural predication and related domains (e.g. the books as opposed to all the books, 20 as opposed to exactly 20). The question arises whether there is reason to believe that the notion of syntactic complexity that matters in this domain is the same as for strengthening.
Prerequisites:
1) Understanding how semantic notions like (contextual) entailment, presupposition, contradiction, etc. are formalized in a possible-worlds framework.
2) A basic understanding of compositional truth-conditional semantics, roughly at the level of the first half of Heim & Kratzer 1998 or a similar textbook (or of a one-week intro semantics course at EGG). Technicalities involving binding etc. will not figure prominently in this class but you should be familiar with the notions of types, function application, truth and definedness conditions, and lambda-terms.
No preparatory reading necessary, but if you have time to read one paper in advance, make it Katzir (2007).
References
Bar-Lev, Moshe E., Itai Bassi & Tue Trinh. 2025. Symmetry breaking, partition by exhaustification, and fatal competition. Manuscript. https://ling.auf.net/lingbuzz/008962
Buccola, Brian, Manuel Križ & Emmanuel Chemla. 2022. Conceptual alternatives. Competition in language and beyond. Linguistics and Philosophy 45. 265–291.
Feinmann, Diego. 2025. On breaking symmetry by complexity. Snippets 47. 8-10.
Fox, Danny & Roni Katzir. 2011. On the characterization of alternatives. Natural Language Semantics 19. 87–107.
Gazdar, Gerald. 1979. Pragmatics: Implicature, presupposition, and logical form. New York: Academic Press.
Haslinger, Nina. 2024. Pragmatic constraints on imprecision and homogeneity: University of Göttingen dissertation.
Haslinger, Nina & Viola Schmitt. 2025. Revisiting the role of structural complexity in symmetry breaking. To appear in: Proceedings of Sinn und Bedeutung 29.
Heim, Irene & Angelika Kratzer. 1998. Semantics in Generative Grammar. Blackwell.
Horn, Laurence R. 1972. On the Semantic Properties of Logical Operators in English. UCLA dissertation.
Horn, Laurence R. 1989. A natural history of negation. University of Chicago Press.
Katzir, Roni. 2007. Structurally-defined alternatives. Linguistics and Philosophy 30. 669–690.
Meyer, Marie-Christine. 2013. Ignorance and Grammar: MIT dissertation.
Percus, Orin. 2006. Antipresuppositions. In A. Ueyama (ed.), Theoretical and Empirical Studies of Reference and Anaphora: Toward the establishment of generative grammar as an empirical science, 52–73. Japan Society for the Promotion of Science.
Sauerland, Uli. 2008a. Implicated presuppositions. In Anita Steube (ed.), The discourse potential of underspecified structures, 581–600. New York: de Gruyter.
Sauerland, Uli. 2008b. On the semantic markedness of phi-features, In Phi Theory. Phi-Features across Modules and Interfaces, eds. Harbour, D., D. Adger, and S. Béjar, 57–82. Oxford/New York: Oxford University Press.
Schwarz, Bernhard & Michael Wagner. 2024a. Symmetry resolution and blocking. In Baumann, Geraldine, Daniel Gutzmann, Jonas Koopman, Kristina Liefke, Agata Renans & Tatjana Scheffler (eds.). Proceedings of Sinn und Bedeutung 28, 829–847. University of Konstanz.
Schwarz, Bernhard & Michael Wagner. 2024b. Resolving symmetry without constraining alternatives. Manuscript. https://ling.auf.net/lingbuzz/008588
Solt, Stephanie. 2018. Approximators as a case study of attenuating polarity items. In Sherry Hucklebridge & Max Nelson (eds.), Proceedings of NELS 48, vol. 3, 91–104.