The Study of Bound Anaphora in the Generative Research tradition

Martin Everaert
Utrecht institute of Linguistics OTS

	NOTE 1: What you will read below is not an article but an informal presentation in written 
form. References are not complete. My English has not been corrected. Hopefully all the 
non-Dutch examples are correct. I have formulated more loosely than I would do if it were 
an article. Keep that in mind while reading it. 
	NOTE 2: The title of this talk/paper suggests a balanced overview of the literature. 
However, the text below is strongly biased by my Germanic view of the world and leaves 
important issues undiscussed. 

Discussing anaphora within the generative research tradition should start with the Binding Theory 
(BT) as formulated in Chomsky's Lectures of Government Binding (LGB) (cf. 1-3)):

(1)	Binding Theory (LGB 188)
 	a. An anaphor is bound in its Governing Category
	b. A pronominal is free in its Governing Category
(2)	" is the governing category for 8 iff " is the minimal category containing 8, a governor of 
8, and a SUBJECT accessible to "
(3)	a.	" is bound by 8 iff " and 8 are coindexed, 8 c-commands " (and 8    is in an 
	b.	" is free iff it is not bound
   	c.	" c-commands 8 iff the minimal maximal projection dominating "       dominates 
8, and " does not dominate 8

The BT as formulated in (1-3) has been very influential in generative research. It is also true that, 
essentially, it hasn't changed much since its introduction. In Chapter 1 I will first discuss what I 
take to be the most central features of the LGB version of the Binding Theory and how they are 
rooted in earlier work. In Chapter 2 I will discuss some important issues that have been raised 
within the context of the study of (bound) anaphora. In Chapter 3 I will give an outline of the 
Reflexivity framework of Reinhart & Reuland, a radical departure of the BT sketched in (1-3). In 
the final chapter I will discuss some issues that, in my mind, deserve more attention.

Chapter 1 

1.1 BT is not only about reflexives, reciprocals and pronouns

One might be tempted to think that BT only deals with the distribution of reflexives, reciprocals and 
pronouns. That is not the case. From Chomsky' Conditions on Transformations (1973) onwards, 
the study of reciprocals, reflexives and pronouns was always tied to the study of other grammatical 
phenomena such as WH-movement, NP-movement, or control. 
	In Chomsky (1973) two major restrictions on rules of extraction/insertion are introduced, 
the Specified Subject Condition and the Tensed-S condition. He observes that both reciprocal 
formation and WH-movement appear to be subject to the Specified Subject Condition (SSC), as 
the b-examples of (4-5) make clear:

(4)	a.	The meni saw pictures of each otheri
	b.	*The meni saw John's pictures of each otheri
(5)	a.	Whoi did you see pictures of ei
	b.	*Whoi did you see John's pictures of ei

In Chomsky (1980) the Tensed-S condition is replaced by the Nominative Island Constraint (NIC):

(6)	A nominative anaphor cannot be free in S'

The NIC not only excludes examples as in (7a), but also instances of NP-movement as in (7b), 
under the assumption that NP-traces are anaphors subject to condition A, and even instances of 
WH-movement, under the assumption that WH-traces are (A'-)anaphors. 

(7)	a.	*Johni expected that himselfi would win
	b.	*Johni seems that ei would win
	c.	*Whoi do you think [ ei that ei saw Bill] 

This parallelism between lexical anaphors and empty categories was lost when the NIC was 
replaced by the Empty Category Principle. However, Kayne (1984) argues that the Connectedness 
Condition, his alternative to the ECP and devised for WH-movement dependencies, might also hold 
for anaphor antecedent dependencies. 
	Observe that BT generalizes over empty categories and lexical categories. Lexical elements 
are partitioned by means of two features, <" Anaphor> and <" Pronominal>, resulting in the cross- 
classification in (8a); (8b) presents lexical and non-lexical instantiations: 

(8)	a.	<+A,-P>	b.	reflexives, reciprocals, NP-trace
      	  	<+A,+P>  	      	PRO
	        	<-A,-P>		R-expressions, WH-trace
        		<-A,+P>     		pronouns, pro 

The precise relationship between WH-movement, NP-movement, or Control and the binder-bindee 
relationship is, of course, more complicated than I am able to discuss here. But the crucial 
observation is that in the LGB/P&P framework the distribution of bound anaphora was always 
closely tied to other syntactic phenomena. It was for that reason that binding was part of the 
computational system, to use the terminology of Chomsky (1995).        

1.2 There is only one concept 'anaphor' or 'pronoun'

Note that in the BT, there is only one concept >anaphor, subsuming reflexives and reciprocals. If 
we look at the predecessors of BT, you can see that this neglect of the difference between 
reflexives and reciprocals is an innovation. In the work preceding LGB, Chomsky mainly discusses 
the distribution of the reciprocal each other which he takes to be derived from each-movement, i.e. 
the men hated each other is derived from the men each hated the other(s) (Dougherty 1970). For 
reflexives he follows the analysis of Helke (1970) who puts reflexives on a par with bound 
anaphora in certain idioms such as his way, his cool. This approach is most clearly visible in 
Chomsky (1977), where he formulates the following rules:

(9)	a.	Reciprocal rule: assign to each other the feature [+ anaphoric to i] in a structure 
containing NPi
	b.	Bound anaphora: assign to a pronoun the feature [+ anaphoric to i] in a structure 
containing NPi, in the context [NP - Possessive - Nx]

Chomsky recognizes the morphological differences between the two types of anaphors, but stresses 
their similarities: these rules of construal are subject to the same restrictions such as SSC, TSC. 
	Research in the 80's showed that we should distinguish different types of reflexives and that 
their domains of interpretation differ. Reinhart & Reuland (1991/93), for instance, distinguished 
SE-anaphors from SELF-anaphors and showed how they obeyed different constraints. In Safir 
(1996) and Anagnostopoulou & Everaert (1995) it is, furthermore, shown that the internal syntax 
and semantics of reflexives deserve more attention because it influences their distributional 

1.3 Anaphors and pronominals are in complementary distribution

One of the basic insights with respect to anaphoric binding within Generative Grammar was first 
formulated in Jackendoff (1972: 136) as:

(10)	NP1 is " coreferent with NP2 <" refl> if each of the following conditions holds...      
That is, pronouns (free anaphors) and reflexives (bound anaphors) are in complementary 
distribution. Within the Government-Binding framework of Chomsky (1981) this generalization 
appears unaltered in the Binding Theory: an anaphor must be bound within a specific domain (= 
governing category) while a pronominal must be free in that same domain. 
	Huang (1982) argued, however, that this generalization ought to be abandoned and that 
distinct domains of interpretation ought to be assigned to bound anaphors and free anaphors. This 
position is further developed in Chomsky (1986) on the basis of examples such as in (10):

(11)	a.	Theyi saw [each otheri's friends]
	b.	Theyi saw [theiri friends]

This is one of the most salient cases of non-complementarity of anaphors and prononimals in 
English. To account for this non-complementarity Chomsky introduces the notions BT-compatible 
(12) and Complete Functional Complex (13) which, roughly speaking, replace the notions 
Accessible subject and governing category. 

(12)	Complete Functional Complex (CFC)
	8 is a CFC if all the grammatical functions compatible with a head dominated by 8 are 
contained in 8
(13)	BT-compatible
	Indexing I is BT-compatible with (",8) if:
	(i) " is an anaphor and is bound in 8 under I
	(ii) " is a pronominal and is free in 8 under I

On the basis of these notions, the binding conditions are now redefined as in (14): 

(14)	a.	For some domain 8 and an anaphor ", " must be bound in 8, where 8 is the least 
CFC containing y_ for which there exists an indexing I BT-compatible with (",8)
	b.	For some domain 8 and a pronominal ", " must be free in 8, where 8 is the least 
CFC containing y_ for which there exists an indexing I BT- compatible with (",8)

Let us first take a look at (11a). Both the noun phrase containing the reciprocal and the root 
sentence are complete functional complexes, but only for the whole sentence do we find a BT-
compatible indexing with an antecedent. For the possessive pronoun in (11b) this reversed. It is 
precisely the noun phrase CFC that satisfies the requirement that the pronominal has a BT-
compatible indexing. It means that the binding domain of an element depends on its status as an 

Chapter 2

2.1 Parametric variation

In standard BT there is no other way of describing that languages differ with respect to binding 
phenomena than by either assuming that the anaphors itself have different properties, an option I 
will discuss below, or formulating parameters as in (15) from Manzini & Wexler (1987):

(15) 	a.	" is a governing category for 8 iff " is the minimal category which contains " and 
	b.	PARAMETER-values: has (i) a subject, (ii) an INFL, (iii) a Tense, (iv) an indicati-
ve Tense, (v) a root Tense

Examples that illustrate the parameter choices in (15b) are given in (16):  

(16)	a.	*Johni made me talk about himselfi                  			(i: English)
	b.	Jani hoorde mij over zichi praten    		     		(ii: Dutch)
		John heard me about himself talk
     	c.	*Jani vroeg ons  voor zichi te werken	(ii: Dutch)
		John asked us to help himself'
 	d.	Joni bad oss hjelpe segi                         			(iii: Norwegian)
       		John asked us to help himself
		`John asked us to help him'
     	e.	*Joni ville at jeg skulle snakke om segi
       		John wanted that I should speak about himself
 	f.	J\ni segir ap MarRa elski sigi                			(iv: Icelandic)
       		John says that Maria lovessubj himself
		`John says that Maria loves him' 
     	g.	*J\ni veit ap MarRa elskar sigi
       		John knows that Maria lovesind himself
 	h.	Bill-wai [ John-ga zibuni-0 seme-ta to] omot-ta'  	       (v: Japanese)
     		`Bill thought that John blamed him'

These parameter choices obey the subset condition, which means that they are embedded in well-
articulated theory of learnability. However, it is not difficult to show that the parameter choices 
have empirical problems (see chapter 5). More importantly these parameter settings are a good 
illustration that parametrization within BT is pretty hard to arrive at. It is hard to see how the 
parameter choices form a conceptually coherent set. Furthermore, they are completely unrelated to 
the choice of language. As it is formulated in (15), it is pure chance that Icelandic takes indicative 
tense as an 'accessible subject' while Dutch and German take an INFL node. 
	Another line of research focused on the classification of anaphoric elements. The question 
is whether the classification of pronominal elements in BT is subject to parametric variation? Does 
the language learner have to make a choice between, for instance, non-pronominal anaphors vs. 
pronominal anaphors, connectedness anaphors vs. containment anaphors (Hellan 1988), or normal 
anaphors vs. long distance anaphors (Anderson 1986, Koster 1987, Yang 1983)? This immediately 
triggered the question whether we have to distinguish several types of reflexivization? I will come 
that point in the next section. 
	It has frequently been observed that only specific anaphors participate in Long-distance 
binding (cf. Yang 1983). In the Germanic languages only simple reflexives - elements like Dutch 
zich, Norwegian seg and Icelandic sig - and reflexive possessives - such as Norwegian sin and 
Icelandic sinn/sRna/etc. - permit long distance binding; reciprocals - Dutch elkaar, Norwegian 
hverandre - and complex reflexives - Dutch zichzelf, Norwegian seg selv and Icelandic sj and <-LD anaphor> respectively. Although 
descriptively adequate, it clearly would offer no insight and amore principled reason is needed to 
explain why the facts are as they are. Or to put it differently, why is LD limited to a specific class of 
anaphors (cf. Pica 1987)?
	There seems little disagreement that complex anaphors have to be locally bound, clause-
bounded, and that simplex anaphors generally allow non-clause bounded binding. The question is 
how this derived. Several suggestions have been made in the literature, Katada (1991), Hestvik 
(1992), and Progovac (1993), among others, which I will not discuss here.

2.2 Types of reflexivization

In discussing reflexivization facts from Icelandic, ThrExperiencer>Goal/Source/Location/Benefactor>Theme  

Pollard & Sag admit that there is not much empirical difference between (42a) and o-
command. They give two examples which might seem problematic for a thematic 

(43)	a.	Johni was given a book by himselfi	
	b.	I sold the slavei to himselfi

Dalrymple (1993) shows how a case like (11a) might follow from a thematic approach. 
There is a long tradition of accounts which tries to reduce the c-command requirement 
of the binding conditions to a thematic hierarchy constraint, either directly or indirectly. 
As far as we can see such analyses are not less successful than the c-command 
approaches (cf. Anagnostopoulou & Everaert (1997)).  

4.2 Configurationally-sensitive anaphoric dependencies

4.2.1 The [+ N] distinction

The examples below illustrate that both Dutch (44a) and Norwegian (44b) allow long 
distance binding in bare infinitives:

(44)	a.	Jani hoorde mij over zichi praten	
		John heard me about himself talk
  	b. 	Joni hrrte meg snakke om segi	
		Jon heard me talk about himself
  		`Jon heard me talk about him'	

If we assume a BT which incorporates LF-movement of anaphors, the grammaticality of 
(29) can be accounted for if we can show that the SE-anaphor is able to attach to the 
matrix INFL through LF-movement. In an analysis along the lines of R&R the SE-
anaphor first moves to the embedded verb, then to the embedded INFL-node and, 
consequently, to the matrix verb and INFL-nodes, giving the desired reading. 
 	The examples in (45) show that long distance binding in adjectival small clauses 
is ungrammatical (cf. Everaert 1986, Broekhuis 1994; the German example (45c) is from 
Reis 1976):

(45)	a.	??dat hiji mij jaloers op zichi maakte	 
       		that he  me jealous of him makes
		`that he made me jealous of him'
	b.	??dat hiji mij te snel tevreden over zichi vond
		that he  me too quickly satisfied with himself found
		`that he considered me too quickly satisfied with him'
     	c.	*Hansi lăsst den Vater ruhig stolz auf sichi sein	
		Hans let the father easily proud of himself be
		`Hans let the father easily be proud of him'

Given the analysis we have just sketched for (29) this is surprising. There is no a priori 
reason to assume that LF-movement in adjectival small clauses behaves differently from 
verbal small clauses. To make the picture even more complicated observe that the same 
type of configuration does not block long distance binding in Norwegian (46a) or 
Icelandic (46b)

(46)	a.	Joni gjorde dem glade i segi	
		John made them fond of himself
		`John made them fond of them'
  	b.	SŠlfrŹȚingurinni gerȚi Harald stoltan af sLri	
		The psychiatrist made Harald proud of himself
		`The psychiatrist made Harald proud of him'

Using theoretically neutral terms, statement (47) seems to correctly describe the facts in 

(47)	a.	Maximal projections of [+N]-heads are binding domains
	b.	Dutch/German chooses (i) as a parameter, the Scandinavian languages do 

Given such a generalization and parameter setting, we would correctly predict that NPs 
are opaque domains for binding in Dutch (48), but not in languages like Norwegian 
(49a) and Icelandic (49b,c) (Everaert (1986,) Broekhuis (1994)):

(48)	a.	*Jani ontmoette enige vrienden van zichi  
		`Jon met some friends of himself'
	b.	*Jani zag Marie's foto van zichi
		`John saw Mary's picture of himself'
(49)	a.	Jon traff noen venner av seg	
		`Jon met some friends of himself'
  	b.	JÖni  heyrȚi lysinga af sLri	
		`John heard a description of himself'
  	c.	JÖni heyrȚi lysinga Mar-u af sLri	
		John heard Maria's description of himself
		`John heard Maria's description of him'

Clearly it needs to be explained why (47) should hold and how it follows from a 
movement analysis. Still, these facts seem to indicate that binding is configurationally 

4.2 The NP/PP-object distinction

(44) illustrated that bare infinitives in Dutch and Norwegian are transparent for binding. 
However, this is only part of the story. Compare (44) with the examples in (50) (the 
Dutch example (50a) from Everaert (1986,) the German example (50b) from Reis (1976), 
the Danish example (50c) from Vikner (1985)):

(50)	a.	*Michaeli liet Peter zichi overtuigen	
		Michael let Peter himself persuade
		`Michael let Peter persuade him'
     	b.	*Hinzi lăsst Kunz sichi den Brief geben	
		Hinz let Kunz himself the letter give
		`Hinz let Kunz give him the letter'
  	c.	Michaeli lod Peter overtale sigi	
		Michael let Peter persuade himself
		`Michael let Peter persuade him'

Apparently long-distance binding from bare infinitives is allowed only in the case of PP-
objects in Dutch and German and not in the case of DP objects. The Scandinavian 
languages behave differently in this respect as (44b) and (50c) show. 

The statement in (51) seems to correctly describe the facts in (44) and (50): 

(51)		(i)		Bare infinitives are binding domains for verb-governed anaphors
 		(ii)		Dutch/German chooses (i) as a parameter, the Scandinavian languages do 

And, again, the facts seem to indicate that binding is configurationally sensitive and the 
question must be raised why (51) should hold and how it follows from a movement 
analysis as sketched. In the context of this paper we will not be able to address these 

5 Conclusions
In this paper I have discussed some of the most important features of the Binding 
Theory in the Government-Binding/Principles and Parameters framework as it has 
developed over the year. I have briefly discussed the Reflexivity framework and and 
tried to show . Finally we have looked at the question to what extent anaphoric 
dependencies could be called `configurationally sensitive'. 


Anagnostopoulou, Elena, and Martin Everaert. 1995. How Exceptionel are Nominative 
anaphors? A Case Study of Greek. In Actes du deuxieme colloque de Langue & 
Grammaire, ed. LJa Nash, George Tsoulas, Anne Zribi-Hertz, University Paris 8, 
Dept. des Sciences du Langage, 19-32
Anagnostopoulou, Elena, and Martin Everaert. 1997. Towards a more complete 
typology of anaphoric expressions, ms Utrecht University & Tilburg 
  Anderson, Stephen, R. 1982. 	Types of Dependency in Anaphors: Icelandic and other 
Reflexives, Journal of Linguistic Research 2: 1-22
    Battistella, Edwin. 1989. Chinese Reflexivization: A Movement to Infl approach. 
Linguistics 27:987-1012
    Broekhuis, H.  1994. The referential properties of noun phrases I. Modern Grammar of 
Dutch, vol. 1. Tilburg University.
    Cantrall, William. 1974. Viewpoint, Reflexives and the Nature of Noun Phrases. The 
Hague: Mouton.
  Chomsky, Noam 1973 Conditions on Transformations, In A Festschrift for Morris Halle.  
Eds. Stephen Anderson and Paul Kiparsky, New York: Holt, Rinehart and Winston: 
  Chomsky, Noam 1976. Conditions on Rules of Grammar, Linguistic Analysis 2: 303-351.
    Chomsky, Noam. 1977. On WH-movement. In Formal syntax. Eds. Peter Culicover at 
al. New York: Acadamic Press. 71-132.
    Chomsky, Noam. 1980. On Binding, Linguistic Inquiry 11, 1-46. 
    Chomsky, Noam. 1981. Lectures on Government and Binding. Dordrecht: Foris.
    Chomsky, Noam. 1986. Knowledge of Language: its nature, origin and use, New York: 
    Chomsky, Noam. 1995. The Minimalist Program. Cambridge, Massachusetts :MIT 
    Chomsky, Noam and Howard Lasnik. 1993. The Theory of Principles and Parameters, 
In Syntax: An International Handbook of Contemporary Research, eds. Joachim Jacobs, Arnim 
von Stechow, Wolfgang Sternefeld and Theo Vennemann, Berlin: De Gruyter. Reprinted 
in Chomsky (1995).
    Cole, Peter and Li-May Sung. 1994. Head Movement and Long-Distance Reflexives. 
Linguistic Inquiry 25, 355-406
    Dougherty, Ray. 1970. A grammar of coordinate conjoined structures, Language 46.4
    Everaert, Martin. 1986. The Syntax of Reflexivization, Dordrecht: Foris Publications.
    Everaert, Martin. 1991 Contextual determination of the anaphor/pronominal 
distinction. In Long-distance anaphora. Eds, Jan Koster, Eric Reuland. Cambridge: 
Cambridge University Press, 77-118.
    Everaert, Martin & Elena Anagnostopoulou. 1997. Thematic Hierarchies and Binding 
Theory: Evidences from Greek. In Empirical Issues in Formal Syntax and Semantics. eds. 
Francis Corblin, DaniPle Godard, Jean-Marie Marandin, Bern: Peter Lang. 43-60.
  Giorgi, Alessandra. 1984.	Toward a Theory of Long Distance Anaphors: a GB 
approach, The Linguistic Review 3: 307-362.
    Helke, Michael. 1979. The grammar of English Reflexives, New York: Garland publications.
    Hellan, Lars. 1988. Anaphora in Norwegian and the Theory of Grammar. Dordrecht: Foris 
    Helke, Michael. 1979. The Grammar of English Reflexives, New York: Garland 
    Hestvik, Arild. 1992. LF-movement of Pronouns and Antisubject Orientation, Linguistic 
Inquiry 23:557-94.
  Huang, James. 1982. Logical Relations in Chinese and the Theory of Grammar, Ph.D. 
disseration, MIT.], Cambridge, Massachusetts
    Huang, C.-T. James, and C.-C. Jane Tang. 1991. The local nature of the long-distance 
reflexive in Chinese. Long-Distance Anaphora. Eds. Jan Koster and Eric Reuland. 
Cambridge: Cambridge University Press. 263-82
    Huang, Yan. 1994. The syntax and pragmatics of anaphora. Cambridge: Cambridge 
University Press.
    Jackendoff 1972 Semantic Interpretation in Generative Grammar, Cambridge, 
Massachusetts: MIT press.
    Jackendoff, Ray 1990. Semantic Structures. Cambridge, Massachusetts :MIT Press.
    Jackendoff, Ray.1992. Mme. Tussaud Meets the Binding Theory. Natural Language and 
Linguistic Theory. 10: 1-32.
    Katada, Fusa. 1991. The LF representation of anaphors. Linguistic Inquiry 22: 287-313.
    Kayne, Richard. 1984. Connectedness and Binary Branching. Dordrecht: Foris 
    Koster, Jan. 1988. Domains and Dynasties. Dordrecht: Foris Publications.
    Kuno. Susumo. 1987. Functional Syntax: anaphora, discourse and emphaty. Chicago: The 
University of Chicago Press.
    Lebeaux, D.  1983. A distributional difference between reciprocals and reflexives. 
Linguistic Inquiry, 14: 723--30 
    Levinson. Stephen. 1991. Pragmatic reduction of the Binding Conditions revisited. 
Journal of Linguistics 25: 445-72
    Lidz, Jeffrey. 1995. Morphological Reflexive Marking: Evidence from Kannada. 
Linguistic Inquiry 26, 705-710.
    Maling, Joan. 	1984. Non-Clause Bounded Reflexives in Icelandic, Linguistics and 
Philosophy 7: 211-241.
    Pica. Pierre. 1987. On the nature of the reflexivization cycle. Proceedings of the North-
Eastern Linguistics Society 17: 483-99.
    Philip, William., and Peter Coopmans. 1996. The Role of referentiality in the 
acquisition of pronominal anaphora. In: K. Kusumoto (ed.) Proceedings of the North East 
Linguistics Society 26, GSLA, pp. 241-255.
    Pollard, Carl., and Ivan Sag. 1994. Head-Driven Phrase Structure Grammar. Stanford: 
CSLI, and Chicago: The University of Chicago Press.
    Reinhart, Tanya., and Eric Reuland. 1991. Anaphors and logophors: an argument 
structure perspective. In Long-Distance Anaphora. eds. Jan Koster and Eric Reuland. 
Cambridge: Cambridge University Press. 283-334.
    Reinhart, Tanya., and Eric Reuland. 1993. Reflexivity. Linguistic Inquiry 24.4: 657-720.
    Reuland, Eric, and Tanya Reinhart. 1995. Pronouns, Anaphors and Case. In Studies in 
Comparative Germanic Syntax, ed. Hubert Haider, Susan Olsen and Sten Vikner. 
Dordrecht: Kluwer Academic Publishers, 241-268.
  Reis, Marga. 1976. Reflexivierung in deutsche A.c.I.-konstruktionen, Ein 
transformations-grammatisches Dilemma, Papiere zur Linguistik 9: 5-82.
    Reuland, Eric. 1996. Pronouns and Features. In: K. Kusumoto (ed.) Proceedings of the 
26th annual meeting of the North Eastern Linguistic Society, University of Massachusetts, 
Amherst, 319-333.
    Reuland, Eric, and Tanya Reinhart. 1995. Pronouns, Anaphors and Case. In Studies in 
Comparative Germanic Syntax. eds. Hubert Haider, Susan Olsen and Sten Vikner. 
Dordrecht: Kluwer Academic Publishers, 241-268.
Roberts, Ian. 1991. Excorporation and Minimality. Linguistic Inquiry 22, 209-218.
Safir, Ken. 1996. Semantic Atoms of Anaphora. Natural Language and Linguistic Theory 14,