Analogy in Language and Linguistics

by

Olga Fischer

• LAST REVIEWED: 29 May 2019
• LAST MODIFIED: 29 May 2019
• DOI: 10.1093/obo/9780199772810-0238

Introduction

Analogy (from Ancient Greek αναλογια, “reasoning from one point to another”) is a cognitive process involving comparison whereby the information concerning one element is linked to another element through observed similarities. Analogy is related to “iconicity” (both involve parallels of form/meaning): the strong drive language users have to describe their world through signs that have some natural connection with the object the sign refers to. Analogy is concerned with (1) similarities between observable properties (material or horizontal analogy) and (2) causal similarities, i.e., the relations between a material property and a function of that property (vertical analogy). It enables language learning, is involved in language change, and helps in seeking explanations in science (including linguistics) through the spotting of parallels, e.g., the discovery of sound waves on the basis of water waves. Analogy is omnipresent in language but its working through speakers in language use and language change can only be captured after the fact; there are no fixed “rules” to “predict” analogy. It plays an important role in streamlining (“oiling”) the “machinery” (i.e., the “rules” or “patterns”) of language: without analogy, languages may develop too many exceptions, caused by internal changes elsewhere in the system or external factors, such as contact. The cognitive ability people have to analogize is the concern of cognitive scientists. In linguistics proper, a distinction is often made between this ability and the mechanism of analogy, which is concerned with the effects it has on language output, with specific patterns and constraints. This more narrow approach bears most fruit on the morpho-phonological level, where formal similarities play an important role. As soon as higher lexical and grammatical levels are involved, similarities become looser, more associative, making it more difficult to constrain the workings of analogy. Some linguists make a distinction between proportional and non-proportional analogy: the former is seen as more systematic, more concerned with form, and the result of interference from a productive morphological system (as in analogical extension/leveling), while the latter concerns innovations in individual items and constructions, with meaning playing a large(r) role. In morphological studies, analogy is usually seen as strictly proportional, making use of the formula A : B :: C : D, with more sporadic forms of analogy excluded as being non-proportional. Linguists interested in analogy on higher levels (syntax, lexis, semantics), also consider analogy to be proportional but in a less strict sense; here too there is symmetry between the forms and functions of source and target. For example, in the metaphorical use of foot, when there is a symmetry in meaning between referent A (“bottom part of a human body”) and B (“bottom part of a mountain”), this may result in a new formal symmetry between the signs: thus C foot leads to the new use of foot in D (just like work: worked leads to help: helped (replacing earlier holp).

General/Foundational Works

Analogy is such a pervasive and important factor in language that it is impossible to do justice to all the studies that have paid attention to it in its various forms. In this overview, a distinction is made between the different areas in which analogy has been discussed, and in each of the sections in this bibliography only a small number of the most relevant works will be included. (Text)books and book and journal articles will all be discussed together. No special journal or series is devoted to analogy. Analogy has a natural place in well-known general linguistic works in the first half of the 20th century (for the main Neogrammarian interest in analogy, see Analogy and Historical Linguistics/Language Change), such as Sapir 1921, Bloomfield 1933, and Hockett 1958. In these works, analogy is mostly restricted to the morphological level. Analogy as a useful concept virtually disappeared in the generative period—when the syntactic module acquired the central role in grammar—since it was felt to be unamenable to fixed rules or principles. Later on, it began to reassert its place in linguistic discussions as a reaction to the growing influence of transformational generative grammar, as can be seen in Best 1973, Esper 1973, and Anttila 1977. Esper shows how important the work on analogy Paul 1909 was: this work not only dealt with analogy from a diachronic point of view as was usual with the Neogrammarians, but also saw its importance for the synchronic study of language. With the rise of cognitive and usage-based approaches to grammar, and the interest in cognitive science in the last quarter of the 20th century, new studies appeared which take analogy as a general starting point, such as Slobin 1985, Itkonen 2005, Wanner 2006, and the articles in Blevins and Blevins 2009. More attention is now also paid to analogy in syntax, where meaning plays a crucial role next to form. Wanner 2006 presents the most elaborate work on analogy, in which the author presents and works out his “soft syntactic” model. In this system, “rules are secondary reorganizations of varying generalization depending on the constitutive force of analogy for their aspect and implementation” (p. 197). Wanner stresses both the difficulty of the individual to spot differences (because the seeing of similarities is filtered by subjective perception) and the incremental nature of learning, which leads to categorization, increasing abstraction, and connections of similarity forming a “network.” It also pays attention to the social dimensions involved in analogy (prestige, frequency, standardization).

• Anttila, Raimo. 1977. Analogy. The Hague: Mouton.

  DOI: 10.1075/lisl.1

  A classic study, which deals with all aspects of analogy then known, such as the classical tradition, works written by French, Italian, German, and Finnish linguists and psycholinguists in the 19th and 20th centuries, semiotics (the theory of signs), the categorization of analogy, its role in language change, the absence of analogy in (diachronic) generative grammar, and the link between analogy and contiguity.

• Best, Karl-Heinz. 1973. Probleme der Analogieforschung. Munich: Hueber.

  One of the earliest volumes to take up the question of analogy again after it had become neglected within the Chomskyan School of linguistics. He gives a useful overview of analogical research since Greek and Roman times, paying special attention to the Neogrammarians (see Analogy and Historical Linguistics/Language Change), and sets out the problems connected with analogy research.

• Blevins, James P., and Juliette Blevins, eds. 2009. Analogy in grammar: Form and acquisition. Oxford: Oxford Univ. Press.

  This volume is the result of the first workshop on analogy at the Max Planck Institute for Evolutionary Anthropology in 2006. The editors stress that the ability for analogical reasoning forms the core of human cognition, a fact already recognized in ancient times, and they argue against Chomsky’s dismissal of analogy. Topics covered in the volume include the role of analogy in learning, in phonology (sound symbolism), morphology (inflexional systems and compounding), grammar (categorization), semantics (metaphor), and analogical modeling.

• Bloomfield, Leonard. 1933. Language. Chicago: Univ. of Chicago Press.

  An early structuralist general book on all aspects of language, which has a separate chapter on “Analogic Change.” This deals mainly with the more usual “proportional analogy” (leveling, extension, backformation, compounding, etc.) at the morpho-phonological and lexical levels. What Bloomfield calls “regular analogy” is dealt with briefly. This refers to pattern building in syntax that enables speakers to utter speech forms they have not heard before on the analogy of similar forms they have heard.

• Esper, Erwin A. 1973. Analogy and association in linguistics and psychology. Athens, GA: Univ. of Georgia Press.

  One of the classics. A very readable psycholinguistically oriented book, which fell into disuse in the generative period. (Generative grammar deeply influenced many psycholinguists at the time, causing them to concentrate on the “reality” of generative “rules” rather than the “slipperiness” of analogy.) The book offers a survey of theories, providing insight into the role of analogy and association in human behavior from classical times up to the present. It discusses definitions for the term “analogy” and presents a number of experimental case studies involving analogy.

• Hockett, Charles F. 1958. A course in modern linguistics. New York: MacMillan.

  DOI: 10.1111/j.1467-1770.1958.tb00870.x

  Hockett emphasizes, like Paul 1909 before him, the role played by analogy in learning. Analogy is further discussed in connection with change. He mentions three mechanisms: (i) sound change, (ii) borrowing, and (iii) analogical creation. He does not mention reanalysis, which is usually seen as the main mechanism of language change in historical linguistic research (see Analogy and Historical Linguistics/Language Change). He also notes some minor mechanisms (contamination, metanalysis, metathesis, haplology, and assimilation/dissimilation) that are all considered to be closely allied to analogy.

• Itkonen, Esa. 2005. Analogy as structure and process: Approaches in linguistics, cognitive psychology and philosophy of science. Amsterdam: Benjamins.

  DOI: 10.1075/hcp.14

  A very broad approach to analogy but with an emphasis on its role in language and linguistics. It begins by explicating the general concept and its limits, the relation to iconicity, its innateness, complexity, and context dependency. It further deals with analogy in various domains or levels of language, in different linguistic models (both synchronic and diachronic), and in oral and signed languages. It also discusses analogy in the context of other cognitive domains, such as iconicity, vision, music, logic, and science in general.

• Paul, Hermann. 1909. Prinzipien der Sprachgeschichte. 4th ed. Halle, Germany: Niemeyer.

  Paul offers a very broad view of analogy, unlike most Neogrammarians. Especially important are the chapters on semantic change and analogy. Analogy (or the recognition of similarity) is considered prevalent not only in morphological change, but also in language learning, syntax, semantics, and the lexicon. His thinking predates works on metaphor theory by scholars like George Lakoff and Mark Johnson, conceptual blending as in the work of Mark Turner, and neuronal networks and the idea of system mapping in cognitive science (see Analogy and Cognitive Science: Language Learning/Processing). First edition published 1880.

• Sapir, Edward. 1921. Language: An introduction to the study of speech. New York: Harcourt Brace.

  Discusses the phenomenon of analogical leveling in his chapters on “Phonetic Law” and “Drift,” where he sets out how a certain pattern in a language may slowly increase and then continue to influence other patterns that show some similarity to it. This is basically an analogical process on a fairly abstract level of the grammatical system. The idea of drift was later developed by typologists such as Greenberg to refer to word order symmetries, and it also surfaces in generative grammar in the form of X-bar theory.

• Slobin, Dan I., ed. 1985. The crosslinguistic study of language acquisition. Vol. 2, Theoretical issues. Mahwah, NJ: Erlbaum Associates.

  Introduces the idea of “operating principles” (article by A. M. Peters) and the learning device or LAD ‘Language Acquisition Device’ (article by D. I. Slobin). These describe and explain developmental changes taking place in children. Rather than innate principles, children are said to rely on developmental “bootstrapping,” i.e., they “pull themselves up” from what they already know (i.e., starting with tokens they hear), and by doing so, get to a higher, more abstract stage and thus create more resources by which to pull themselves up even further. All this is based on their analogical capacities.

• Wanner, Dieter. 2006. The power of analogy: An essay on historical linguistics. Berlin: Mouton de Gruyter.

  DOI: 10.1515/9783110919813

  This study goes beyond historical linguistics since it questions “the quality of explanation afforded by an approach that sidelines the diachronic phenomena characterizing language” (p. 5). To offer more than the various frameworks current in (historical) linguistics in terms of explanation, Wanner postulates a “soft” syntactic model of generalized analogy, based on the cognitive abilities of individual language learners. It is a system without formal/fixed rules, with a network of lexical elements and syntactic chunks linked to a number of “dimensions” used in processing, such as precedence, cohesion, and agreement.