Rules and Exceptions

Using Exceptions for Empirical Research in Theoretical Linguistics

by Christopher Beedham (Volume editor) Warwick Danks (Volume editor) Ether Soselia (Volume editor)
Edited Collection XII, 277 Pages


This book assembles a collection of papers first presented at the Summer School and Conference on the Method of Lexical Exceptions held at the University of St Andrews, Scotland, 2-8 September 2007, which explored an area of linguistics now referred to as ‘the method of exceptions and their correlations’.
Recognition of the work of Saussure was impeded during his lifetime by the Junggrammatiker (Neogrammarians) and their view of exceptions, but this book incorporates exceptions into a Saussurean approach. Exceptions to rules are treated here not as something wilful and inexplicable, but as a clue to what has gone wrong in the original rule.
The topics covered are the passive, irregular verbs, morphology, transitivity, light verb constructions, resultative verbs, compound nouns, phonology, colour terms, historical-comparative reconstruction, language teaching, Saussurean structuralism and the approach of the Junggrammatiker to exceptions. The languages addressed are English, Arabic, Georgian, Turkish, Russian, the Cushitic languages and German. Grammar and linguistics are usually thought of as purely theoretical disciplines, but this book demonstrates how to use exceptions to conduct ‘experiments’ in the manner of the natural sciences, which leads empirically to better theory.

Table Of Contents

  • Cover
  • Title
  • Copyright
  • About the author
  • About the book
  • This eBook can be cited
  • Contents
  • List of Figures
  • List of Tables
  • Preface
  • 1. Exceptions and their Correlations: A Methodology for Research in Grammar
  • Introduction
  • Theoretical background
  • The passive and non-passivizable transitive verbs
  • Identifying the exceptions
  • Conducting ‘experiments’
  • Interpreting the data
  • The Russian passive
  • Recognizing errors and correcting them
  • The measurement of pro and contra examples
  • Time-scale
  • The link with language teaching
  • Tense formation and irregular verbs
  • Working with informants
  • Identifying the exceptions
  • Conducting experiments
  • VCs of irregular and regular verbs
  • Irregular verb VCs and function words
  • Working with research assistants
  • Irregular verb VCs and prefixes in German
  • Conclusion
  • Bibliography
  • 2. Evaluation and Adaptation: Applying the Method of Exceptions and their Correlations to Modern Standard Arabic
  • Introduction
  • The specific challenge of MSA
  • Is the Arabic verb really this regular?
  • How are the verb patterns distributed in the lexicon – is it random?
  • To what extent are the syntactic and semantic properties of verbs predictable from their morphology? Does formal similarity imply semantic similarity?
  • Foundations of the method of exceptions and their correlations
  • From form to meaning
  • Synchronicity
  • Unexplained exceptions are indicative of incorrect analysis
  • The Hegelian triad and scientific method
  • Adapting the method of exceptions and their correlations
  • Revisiting the data
  • Developing analytical tools
  • Lateral thinking and an intuitive leap
  • Reiteration
  • Corpus data and examples
  • Adequacy of methodology
  • Summary
  • Bibliography
  • Online resource
  • Discussion (Extract)
  • Reference to Discussion
  • 3. Rules and Exceptions: Neogrammarians and the Lexicon
  • Introduction
  • The Junggrammatiker and the historical background: Comparative linguistics
  • The principles
  • The results
  • The method
  • Correspondence vs likeness
  • A rationale
  • The Junggrammatiker: A radical criticism of the peaceful world of comparative linguistics
  • Creed
  • Rules
  • Rules and exceptions: The case of Grimm’s law
  • Grimm’s law
  • A drag chain
  • Motivation
  • Exceptions?
  • Cases of unexpected voicing
  • Clusters
  • Residues explained by laws proper to Gothic or Sanskrit
  • Chronology of rules
  • An alternative theory of sound change: Lexical diffusion
  • Lexical diffusion
  • Definition
  • Data from English
  • Data from Chinese dialects
  • Data from Dravidian languages
  • Kiparsky (1995): Lexical diffusion and Lexical Phonology
  • The framework: Lexical Phonology
  • Lexical diffusion as analogy
  • A rebuttal of Kiparsky: Hale 2003
  • Methodological flaws
  • Diachrony
  • Concluding remarks
  • Bibliography
  • Discussion (Extract)
  • Reference to Discussion
  • 4. Passive in Georgian
  • Bibliography
  • Discussion (Extract)
  • References to Discussion
  • 5. On the Specification of Basic Colour Terms in Georgian
  • Bibliography
  • 6. On Compounds in Georgian of the Type cxenip’aria ‘horse stealer’
  • Bibliography
  • Discussion (Extract)
  • 7. A Cognitive Approach to Exceptional Ditransitive Verb Forms in Georgian
  • Introduction: The structure of the Georgian verb
  • Georgian preverbs
  • The conceptual representation of space structuring in Georgian
  • The main dimensions of space structuring and their various combinations
  • Preverbs and exceptional ditransitive verb forms in Georgian
  • The cognitive interpretation of semantic roles: Conceptual structures of EDV
  • Bibliography
  • Discussion (Extract)
  • References to Discussion
  • 8. Грузинский меdиоактивный гᴧагоᴧ
  • Бибᴧиография
  • 9. Lexical Exceptions in the Comparative Reconstruction of the Kartvelian Languages: Words for ‘oak’
  • Bibliography
  • Discussion (Extract)
  • 10. The Non-Suffixal Derivation of Intensive Forms in Turkish
  • Bibliography
  • Appendix
  • Discussion (Extract)
  • 11. On the Discourse Functions and Contrastive Phraseology of Equivalent Light Verb Constructions Involving ‘make’ and ‘take’
  • Introduction
  • LVCs in mainstream grammar
  • LVCs and aspect
  • The syntax of LVCs
  • LVCs in Systemic Functional Grammar
  • LVCs and Transitivity
  • LVCs and Mood
  • LVCs in the British National Corpus
  • The contrastive phraseology of to decide, make + decision and take + decision
  • The contrastive phraseology of to note, make + note and take + note
  • The contrastive phraseology of make + effort / take + effort
  • The contrastive phraseology of make + time and take + time
  • Conclusion
  • Bibliography
  • Discussion (Extract)
  • References to Discussion
  • 12. On a Class of Resultatives in English, with Evidence from Electronic Corpora
  • Introduction
  • Diachronic data and construction grammar
  • The Transitive into -ing pattern in present-day English
  • Summary
  • Bibliography
  • Discussion (Extract)
  • 13. Prefix Verbs in Cushitic are not Exceptions
  • Introduction
  • The weak and the strong inflections
  • Lexical vowels of verbs in Afar
  • Lexical vowels of verbs in Bedja
  • Initial CV, proper government and affix ordering
  • Conclusion
  • Bibliography
  • Notes on Contributors
  • Index
  • Series index

← vi | vii → Figures

Figure 2.1Selected data for co-occurrences of verbal stems.

Figure 3.1Grimm’s law as a drag chain phenomenon.

Figure 3.2Underspecification, default rule application and regularization in Kiparsky 1995.

Figure 7.1A conceptual model of simple preverbs.

Figure 7.2Complex preverb with mo-: Ego Space includes Alter Space. ← vii | viii →

← viii | ix → Tables

Table 1.1The voice analysis versus the aspect analysis of the passive (in English be + V-ed)

Table 1.2A proposed schedule of phases for using the method of exceptions and correlations in a PhD

Table 1.3The basic morphology of English irregular verbs

Table 1.4The basic morphology of German irregular verbs

Table 1.5The basic morphology of Russian irregular verbs ending in –at’

Table 1.6Numbers of irregular verbs and structurally comparable regular verbs in English, German and Russian

Table 2.1Verbal patterns and lexical frequencies

Table 2.2Extract from data tabulated by root and pattern

Table 2.3Co-occurrence of Patterns III and VI

Table 2.4Selected forms with C1āC2 or C2āC3 sequence

Table 3.1Correspondences between Gothic and Sanskrit, Greek and Latin

Table 3.2Two in Gothic and Sanskrit, Greek, Latin and Armenian

Table 3.3Correspondences between Indo-European, Greek and Armenian

Table 3.4Old High German consonant shift

Table 3.5Correspondences between Gothic, Sanskrit, Greek and Latin

← ix | x → Table 5.1Results of list test: Frequency of terms

Table 5.2Results of list test: Average place in the list

Table 5.3Results of list test: Percentage score within a place number

Table 10.1Prefix-like morphemes (interfixes) in Turkish

Table 11.1Light Verb Constructions and verbal group complexes

Table 11.2The distribution of make and take in fifty different Light Verb Constructions

Table 12.1Semantic classes of verbs selecting the transitive intoing pattern in the third part of the Corpus of Late Modern English Texts

Table 12.2Matrix verbs selecting the transitive intoing pattern in the British Books Corpus of the Bank of English

← x | xi → Preface

The papers in this volume are revised and updated versions of papers delivered at the Summer School and Conference on the Method of Lexical Exceptions held from 2 to 8 September 2007 at the University of St Andrews, Scotland, organized by myself and Warwick Danks.1 From the publication of this volume the above mentioned method is now called the method of exceptions and their correlations, for reasons which are given in my paper. Appended to each paper is an extract from the discussion which followed the paper, as transcribed at the time, with updates where necessary.

I am especially grateful to Prof. Thomas V. Gramkrelidze, University of Tbilisi, Georgia, for recommending the Summer School and Conference to his colleagues and students and facilitating their attendance, and for giving permission to publish here slightly revised versions of two papers first published in the Bulletin of the Georgian National Academy of Sciences. The contribution of the Georgian participants was invaluable, both quantitatively and qualitatively. I am grateful to all those who attended and gave papers at the Summer School and Conference, making it such an interesting and fruitful event. Thanks also to my colleagues, students and the secretarial staff in the School of Modern Languages and in SAILLS (St Andrews Institute of Language and Linguistic Studies) for their help, advice and support. I am grateful also to the Russell Trust and to the Carnegie Trust for the Universities of Scotland for their financial support.

Christopher Beedham

St Andrews

21 November 2013 ← xi | xii →



1See <http://www.st-andrews.ac.uk/modlangs/research/conferences/pastevents/>, accessed 21 November 2013.


1Exceptions and their Correlations: A Methodology for Research in Grammar


A method of empirical research is described in which unexplained exceptions to a rule and their correlations are used in a systematic way to lead the researcher to a revised version of the rule which explains and removes the rule’s anomalies, especially the exceptions which one started out with. Two rules and their exceptions in English, German, and Russian are presented as case studies in the method: the passive and non-passivizable transitive verbs; tense formation and irregular verbs. It is hoped that other linguists will try out the method on their own chosen constructions in their own languages.


In this paper we will look at how unexplained exceptions to a rule and their correlations can be used in a systematic way to lead the researcher to a revised version of the rule which explains and removes the rule’s anomalies, especially the exceptions which one started out with.1 We will examine two constructions and their exceptions as case studies in the method – the ← 1 | 2 → rule of passive and non-passivizable transitive verbs, tense formation and irregular verbs – in English, German and Russian. The examination of these two areas will be brief, because the paper is about method, not about the passive or irregular verbs, but references will be given to enable the reader to follow up the analyses presented. Because the reason for writing the paper is to encourage other linguists to try out the method on their own chosen constructions in their own languages some detail of the history of the method, how it developed over time, how long it took to carry out the data collection, is included in order to give the reader an idea of the timescale involved if he or she decides to try it out. For the same reason some further practical details about using reference works, native speaker informants, and research assistants are also included. (For further details on all these matters – the method, the passive, and irregular verbs – see Beedham 2005b; Danks 2011, this volume. On exceptions in general see Simon and Wiese (eds) 2011, Corbin and Dessaux-Berthonneau 1985.)2

Theoretical background

The approach taken by the method of exceptions and their correlations is descriptive (as opposed to generative) and based on Saussurean structuralism (Saussure 1983). The term ‘descriptive’ is used here despite the fact that it is misleading in that some linguists claim that descriptive linguistics merely describes, without explaining. However, in our view ‘descriptive’ linguistics is indeed analytical – it produces analyses – and explanatory, i.e. ← 2 | 3 → theoretical (in the pre-Chomskyan sense of ‘theoretical’): the explanation is in the description. There is a link with pedagogical grammars and language teaching: the areas of grammar and lexis which descriptive grammarians choose to examine are usually the ones which foreign learners of a language find difficult. And for descriptivists the ultimate test of a new theoretical analysis is whether it is taken up in pedagogical grammars.

It is useful though not essential when using the method of exceptions and their correlations to investigate the same formal construction in two or more languages, because it gives you two or more different angles on the same construction. It is also helpful if one of your languages is a foreign language, i.e. not your native language, because you then automatically have the psychic distance necessary to carry out an objective, scientific analysis. It is all too easy when analysing one’s native language to fall into the trap of thinking that the categories and rules one sees there are natural and logical and based on the way the world is, rather than the language-specific, idiosyncratic and arbitrary (in the Saussurean sense) categories and rules which they, in fact, are. You should be an advanced learner of the foreign language in question and hence speak it reasonably fluently: this is necessary to enable you to adjudicate the grammaticality judgements of your native-speaker informants. On the other hand, if one of your languages is your native language that is also handy because of the greater intuitive insights you have for that language. For both foreign language and native language the researcher is actually trying to bring out his or her native speaker(-like) intuitions into an explicit, scientific analysis. Having said it is best to investigate your chosen construction in two or more languages, of course the arguments in support of an analysis in a given language have to come from the structure of that language, one cannot argue across languages and say, for example, because the situation in Russian is this I want to analyse a construction of English like this. The arguments for English have to come from the structure of English.

← 3 | 4 → The passive and non-passivizable transitive verbs

One of the problems we face in theoretical linguistics is how to make the discipline empirical. Some would say it does not need to be empirical, it just is a theoretical discipline, but every theory needs a practice; theory and practice go together. Others would say that linguistics is already empirical, e.g. through the use of computer corpora, and that is true, but there is another, langue-oriented and grammar-based way in which we can make it empirical, and that is through the use of unexplained exceptions to grammatical rules. Let us take non-passivizable transitive verbs as an example. The traditional and still widespread rule of passive says that every transitive verb can form a passive, but it is well known that there are a small number of transitive verbs in English, German and Russian which do not form a passive, despite being transitive, e.g. 1 in English, 2 in German and 3 in Russian below (on the passive in English see Quirk et al. 1985: 159–71; in German see Durrell 2002: 307–22, and Helbig and Buscha 1989: 161–88; in Russian see Borras and Christian 1971: 165–73, and Грамматика русского языка 1960: 504–15):


(1)a.James knows Fiona.

b.?Fiona is known by James.

(2)a.Er mag Käse.
‘he likes cheese’


XII, 277
ISBN (Softcover)
Publication date
2014 (April)
Oxford, Bern, Berlin, Bruxelles, Frankfurt am Main, New York, Wien, 2014. 277 pp., 5 fig., num. tables

Biographical notes

Christopher Beedham (Volume editor) Warwick Danks (Volume editor) Ether Soselia (Volume editor)

Christopher Beedham is Lecturer in the Department of German, School of Modern Languages, University of St Andrews. He has published on the passive and irregular verbs in English, German and Russian and is developing a method of research in linguistics which he calls ‘the method of exceptions and their correlations’. Warwick Danks is University Examinations Officer at the University of St Andrews and holds a PhD in Arabic and Linguistics from the University of St Andrews. His PhD thesis on the Arabic verb, in which he used ‘the method of exceptions and their correlations’, was published in 2011 and reviewed in Language in 2012. Ether Soselia is Professor of Linguistics at Tbilisi State University. Her PhD thesis was on the typology of colour term systems and her monograph Semantic Universals and the Kartvelian Languages: Models of Colour Categorization was published in 2009 (in Georgian).


Title: Rules and Exceptions