The Semantics of Derivational Morphology: Theory, Methods, Evidence by Sven Kotowski and Ingo Plag

Sven Kotowski & Ingo Plag
 The semantics of derivational morphology:
 Introduction
 1 Setting the scene
 In some way or other, most authors will agree that derivational morphology is
 characterized by processes that relate particular forms to particular meanings,
 and vice versa. More specifically, all morphology textbooks will tell their read-
ers that a hallmark of derivation is its concept-creating nature, i.e. that derivation
 concerns operations on lexical semantics, which crucially sets derivation apart
 from inflection (see e.g. Haspelmath & Sims 2013: ch.5, Plag 2003:14f.).Tradition-
ally, however, research in the field has been biased in favor of formal aspects over
 semantic aspects. Although derivational semantics has gained some more atten-
tion recently (see e.g. the volumes Arndt-Lappe & Plag 2015; Bauer et al. 2015;
 Rainer et al. 2014), many of the fundamental questions remain under debate. In
 various ways,the articles compiled in this volume contribute to under-researched
 aspects of the semantics of complex words. All contributions have a sound em-
pirical footing, and many of them present innovative methodological approaches
 that allow for formulating and testing new, and more adequate, models of how
 complex words mean.
 In this introduction, we will first present some major questions of derivational
 semantics, and illustrate them with pertinent phenomena from English. Second,
 we will discuss how these problems have been addressed methodologically in
 the past and how more recent methodological developments offer possibilities
 for more adequate treatments, including novel types of semantic representations.
 Third, we will introduce the articles in this volume with as pecial eye on how their
 methodologies contribute to new insights in their different domains of inquiry,
 and how the empirical results speak to theory.

 2 Fundamental problems of derivational
 semantics
 The firs tgeneral issue can be captured under the label of form-meaning mappings
 and comprises three interrelated phenomena: polysemy, affix rivalry, and form-
meaning mismatches. The majority of morphological processes are polysemous
 (see e.g. Lehrer 2003; Rainer 2014), i.e. one form maps onto more than one mean-
ing or reading. A classic example of a polysemous derivational process is the En-
glish suffix-er, which gives rise to various semantic classes of nouns, including
 agents (dancer), locations (diner), instruments (computer), or person nouns indi-
cating residence or origin (Londoner) (cf. e.g. Lieber 2004; Plag 2003). However,
 a fully satisfactory analysis of the polysemy of-er is still pending, and the suffix
 allows for the illustration of problems posed by derivational polysemy in general
 (see Olsen 2020; Rainer 2014). Most broadly, different categories of meaning raise
 the question of whether they ar einfact based on the same underlying meaning, or
 whether they merely share a form by historic accident (see Olsen 2020 on the agent
 and instrument readings of-er), or have developed into distinct affixes. Moreover,
 related to the question of whether affixes are considered linguistics signs, it is
 unclear what is to be taken as polysemous: an affix as such or the derivatives it
 occurs in. In particular, attempts at assigning to an affix a core meaning represen-
tation that is applicable to all semantic patterns is notoriously difficult. As shown
 by Rainer (2014), assuming a highly underspecified meaning for an affix, such as
 ‘person or thing having to do with X’ for-er (see Plag 2003, 89), may cover all at-
tested meaning categories, but oftentimes has to be  sovague that it overgenerates
 massively and predicts categories of meaning not attested for the affix.
 The second facet of form-meaning mapping concerns affix rivalry. This term
 has been applied to cases where different affixes can take the same kinds of item
 as base and overlap regarding the interpretations the resulting derivatives give
 rise to (see Rainer et al. 2019). Or, from another perspective, we find bases that
 can take different affixes to derive sometimes equivalent, sometimes different in-
terpretations of the derived word. Hence, one is faced with a many-to-many map-
ping of form and meaning. Well-known examples concern competition between
 affixes such as-ness and-ity (e.g. opaqueness/opacity; see Arndt-Lappe 2014) or-

ing and conversion nouns (as in cuttingN/cutN; see e.g. Lieber & Plag 2022). Ri
valry can be challenging to account for, as a multitude of factors may have a say
 in the distribution of the rivals (e.g. phonological, semantic, analogical, lexical
 strata). The frequent existence of doublets, sometimes with a meaning difference,
 sometimes without one, complicates matters further.

The Semantics of Derivational Morphology: Theory, Methods, Evidence by Sven Kotowski and Ingo Plag