F i r s t - O r d e r M o d a l Logic T h e o r e m P r o v i n g and Functional Simulation

Andreas N o n n e n g a r t Max-Planck-Institute for Computer Science Im Stadtwald, 6600 Saarbrucken, Germany Email: nonnenga@mpi-sb.mpg.de T e l : (+49) 681 302 5369 Fax: (+49) 681 302 5401 Abstract
We propose a t r a n s l a t i o n approach f r o m m o d a l logics to first-order predicate logic w h i c h c o m bines advantages f r o m b o t h , the ( s t a n d a r d ) rel a t i o n a l t r a n s l a t i o n a n d the ( r a t h e r c o m p a c t ) f u n c t i o n a l t r a n s l a t i o n m e t h o d a n d avoids m a n y of their respective disadvantages ( e x p o n e n t i a l g r o w t h versus e q u a l i t y h a n d l i n g ) . I n p a r t i c u l a r i n the a p p l i c a t i o n t o serial m o d a l logics it allows considerable s i m p l i f i c a t i o n s such t h a t o f t e n even a s i m p l e u n i t clause suffices in order to express the accessibility r e l a t i o n p r o p erties. A l t h o u g h we restrict the approach here to firstorder m o d a l logic t h e o r e m p r o v i n g it has been s h o w n to be of w i d e r interest, as e.g. sorted logic or t e r m i n o l o g i c a l logic.

Introduction

T o d a y ' s c a l c u l i for m o d a l logics can b e d i v i d e d i n t o t w o m a i n groups. T h e f i r s t g r o u p i s concerned w i t h the extension of already e x i s t i n g c a l c u l i for classical logic by s u i t a b l e a d d i t i o n a l inference rules. T y p i c a l examples can be f o u n d in [ F i t t i n g , 1983] where the well k n o w n t a b l e a u calculus a n d t h e sequent calculus are a p p r o p r i ately extended a n d [ W a l l e n , 1987] in w h i c h t h e connect i o n m e t h o d i s u t i l i z e d for reasoning w i t h i n m o d a l logics. T h e o t h e r g r o u p tries t o e x p l o i t the experience a n d progress m a d e in t h e d e v e l o p m e n t of classical predicate logic calculi a n d defines a t r a n s l a t i o n m o r p h i s m f r o m m o d a l logics to classical logic such t h a t calculi w h i c h are k n o w n to be efficient can be u t i l i z e d . T h e simplest one of those c e r t a i n l y is the so-called r e l a t i o n a l t r a n s l a t i o n m e t h o d ( [ M o o r e , 1980]) w h i c h makes the i m p l i c i t m o d e l semantics e x p l i c i t i n t h e t r a n s l a t i o n . I t s disadvantage lies in t h e size of t h e r e s u l t i n g f o r m u l a e w h i c h get expon e n t i a l l y bigger t h a n the o r i g i n a l ones. Based on W a l l e n ' s ideas, Hans J i i r g e n O h i b a c h a n d others ( [ O h l b a c h , 1989; O h i b a c h , 1 9 9 1 ; A u f f r a y a n d E n j a l b e r t , 1992; Farinas del C e r r o a n d H e r z i g , 1988]) developed the f u n c t i o n a l t r a n s l a t i o n approach w h i c h avoids e x p o n e n t i a l g r o w t h o f the t r a n s l a t e d f o r m u l a e . Here t h e n u m b e r of clauses a n d l i t e r a l s inside clauses is e x a c t l y as b i g as if t h e m o d a l o p e r a t o r s w o u l d have been

ignored (although, of course, literals get bigger by the addition of extra arguments). The disadvantage of this approach lies in the handling of the modal theory (as e.g. reflexivity, transitivity, symmetry, directedness) which gets encoded by certain equations and thus requires a rather strong equality handling mechanism. However, some of these properties and their combinations allow for the definition of suitable theory unification procedures, although finding such an algorithm is indeed a non-trivial task. The approach proposed in this paper can be viewed as a mixture between the relational and the functional translation method. It is tried to combine their respective advantages and to avoid their disadvantages if possible. To be more precise, it consists of both, a relational translation which does not result in an exponential growth in the number of clauses and a functional translation which does not require that the modal theory has to be described by (more or less untractable) equations. This paper is organized as follows: Section 2 is a brief repetition of the modal logic syntax and model semantics. In addition it describes the relational translation formally. Section 3 is concerned with the definition of what we call a functional simulator and its application to the result of a relational translation. This result gives rise to some considerable simplifications which are exemplified in section 4. How this result can be extended to varying domains and multiple modalities is shown in section 4.6 and 4.7. Finally, in section 5, we summarize the effect of the approach.

M o d a l Logics

It is out of the scope of this paper to give an overview over modal logics in general. The reader not familiar with modal logics is referred to [Chellas, 1980] and [Hughes and Cresswell, 1968]. Nevertheless, at least the notions of interpretations and satisfiability have to be repeated briefly, since they will be needed in later sections. By a modal logic interpretation we understand a tuple where W is a non-empty set of worlds and is a binary accessibility relation between worlds, is one of these worlds (the actual world), is a variable assignment and is a mapping from worlds to signature structures which consist of a domain and inter-

80

Automated Reasoning

pretations for the given function and predicate symbols. For convenience let us start w i t h a constant domain structure, i.e. we assume that the domains of the respective signature structures are all identical. Thus, if we refer to some domain V we mean the domain common to all signature structures. The case of varying domains will be handled in section 4.6.

The remaining cases are treated by the usual homomorphic extension. The initial call for the translation of an arbitrary formula then simply is This translation indeed behaves as desired We formulize this by the theorem below. For a proof see eg [Moore, 1980; Ohlbach, 1989; Nonnengart, 1992]) T h e o r e m 2.3 Relational translation is sound and complete, i.e. a modal logic formula is satisfiable if and only if Axioms is (predicate logic) satisfiable, where Axioms denotes those formulae which stem from the additional properties of the accessibility relation of the modal logic under consideration (i.e. the accessibility relation properties). The big disadvantage of this relational translation lies in the exponential growth of the resulting formulae such that already fairly simple theorems can hardly be proved because of the enormous search space. In the following section we therefore introduce an alternative translation method which has its origin in the functional translation approach, but is rather a mixture between functional and relational translation.

3
1 he cases f o r the logical connectives and quantifiers are just as in classical first-order predicate logic and are therefore omitted here. The various modal logics known from the literature mainly differ in the properties associated with the respective accessibility relations. The most common accessibility relation properties are reflexivity, symmetry, transitivity, euclidity, seriality, directedness and linearity. A l l of these properties but one (linearity) will be handled in this paper. 2.1 Relational Translation

Functional S i m u l a t i o n

The idea behind relational translation is to make the implicit semantical parts of modal logic explicit in the predicate logic variant of the given modal logic formula. Hence we assume a new sort W distinct from the domain sort D, a new constant L which is supposed to represent the actual (or current) world, a relation symbol R which denotes the accessibility relation and, for every function and every predicate symbol f (P respectively) a new function symbol /' (predicate symbol P') which accepts one more argument than f ( P ) , namely a world (or actually a term representing a world). The following definition describes the formula morphism which accepts a modal logic f o r m u l a a n d a term w (which denotes a world) and results in a firstorder predicate logic formula. It can be viewed as a direct translation from the satisfiability definition 2.1 into classical logic.

Given a binary predicate R it is possible to split R into predicates R 1 , R 2 . . . such that each of the R i denotes a (partially) functional relation. Intuitively this can be done as follows: arrange the pairs which belong to R, in a two-dimensional array such that every column is responsible for R-pairs with identical first element Having completed this procedure the resulting field contains all element pairs of R and each row of this field represents a subrelation of R which is evidently (partially) functional by construction. Thus, instead of (or additionally to) reasoning with R we can reason with the respective subrelations from the above construction. Note that it is not really necessary to consider the so generated elements as relations (or partial functions). Since by construction there are no more of them than denumerably many (provided R is denumerable) we can equally consider a new sort of the same cardinality and a new function symbol which is supposed to denote something like the applyfunction. The following section provides with a formal definition of such functional simulators and some of its most important properties.

Nonnengart

81

T h e o r e m 3.3 Let M be a (serial) modal logic and let be a M-formula. Then

where Axioms denotes the correspondence axioms f o r the modal logic M and denotes the usual predicate logic satisfiablity relation. P r o o f : From Theorem 2.3 we know that

Thus, and because of the fact that for any R there exists a suitable functional simulator, we obviously have that

3.2

Application to the Relational Translation

In the sequel we assume that the modal logic formulae are already transformed into negation normal form, i.e. all implications and equivalences are removed and the negation signs occur solely in front of the atoms. Evidently, any modal logic formula can easily be transformed into an equivalent one which is in negation normal form. For convenience we also assume in the sequel that we are dealing with serial modal logics unless otherwise stated. The case of non-serial modal logics will be broached in section 4.5. D e f i n i t i o n 3.2 ( T h e F o r m u l a M o r p h i s m Let be a modal logic formula in negation normal form.

simply because under the assumption that FR is a functional simulator of R (i.e. the simulator axioms hold), the two formulae and are logically equivalent. Remains to be shown that is in fact superfluous. Written in clause form gets: where the symbol / is new to the whole clause set. As it is known from the area of equality reasoning, it is never necessary to paramodulate into or from variables. Therefore the equality from needs to be applied only from left to right. However, as a simple induction shows, no term of the form produced by the translation contains a universally quantified FR-variable and thus, no application of the equation from left to right is possible. Since the one direction is not possible and the other direction is not necessary we can simply forget about the whole clause and we are done. We have gained quite a lot already: there is no exponential growth in the number of clauses any more (in fact, the number of clauses generated is exactly as big as if we entirely ignored the modal operators as can be shown by a simple induction; although the clauses themselves might get bigger of course) and the price to be paid is merely the addition of a simple unit clause, namely from Nevertheless, we can do even better. Another fairly easy induction shows that the result of the translation of some arbitrary modal logic formula does not contain any positive R-literals. This is remarkable, for it allows to examine the theory clauses which stem from the properties of the underlying accessibility relation independently of the clauses produced by the translation. A n d this indeed leads to some further considerable simplifications.

The notation u:f is used for readability instead of a p p l y ( f , u). It is shorter and reflects the order of the in the original formula. Thus, for instance, the formula gets translated into where bracketing to the left is assumed. According to the functional simulator properties described above we define the two simulator axioms and as:

and W h a t has been gained so far? It is not too hard to see that the new formula morphism behaves as desired provided the two simulator axioms and are added to the clause set. But introduces an equality and one might expect that equations devour all that has been gained so far. Fortunately it is not that bad. As the following theorem shows this simulator axiom is in fact superfluous and thus merely the unit clause from Sim2 has to be added to the clause set.

Simplifications

The main idea behind the following simplifications is as follows: since we know that the translation of modal logic formulae into predicate logic produces clauses which do not contain any positive R-literals, the only possibility where we can have positive R-literals is via the accessibility relation properties, i.e. via Axioms and Call a clause positive in R if it contains no negative R-literal and consider the set of clauses which are positive in R and which are derivable from Axioms and Sim2 by finitely many resolution steps. We call this set generated from Axioms and then. Now it is not very

82

Automated Reasoning

h a r d to see t h a t t h e set generated f r o m A x i o m s a n d S i m 2 suffices as the m o d a l logic t h e o r y since any negative Rl i t e r a l w h i c h occurs i n A x i o m s can o n l y b e resolved w i t h t h e help of A x i o m s i t s e l f ( a n d R(u, u:x) of course). T h i s means t h a t any clauses w h i c h generate the same set w i l l do f o r o u r purposes as w e l l . For m o r e details see [ N o n n e n g a r t , 1992). I n the f o l l o w i n g a p p l i c a t i o n t o some well k n o w n m o d a l logics we w i l l m a k e use of t h i s f a c t . For h i s t o r i c a l reasons we n a m e the various m o d a l logics w i t h the help of a b b r e v i a t i o n s for t h e respective accessibility r e l a t i o n p r o p e r t i e s . Hence D is used for s e r i a l i t y , B for s y m m e t r y , 4 f o r t r a n s i t i v i t y , 5 for e u c l i d i t y , S4 for r e f l e x i v i t y plus t r a n s i t i v i t y a n d S5 for r e f l e x i v i t y plus e u c l i d i t y . 4.1 T h e Logics K D and K D B

u , v ) a n d t h a t K D 5 a n d K D 4 5 are characterized b y a single w o r l d f o l l o w e d by a non-degenerated cluster w i t h t h e o n l y difference t h a t i n the l a t t e r t h e single w o r l d has access to any w o r l d in the f o l l o w i n g cluster whereas in K D 5 t h i s is n o t necessarily the case. Hence, t h e theories for S5 a n d K D 4 5 are described by a single u n i t clause respectively whereas the t h e o r y for K D 5 requires t w o s i m p l e u n i t clauses. 4.4 Other Interesting Properties

R is called directed if any t w o w o r l d s w h i c h have a c o m m o n predecessor in R also have a c o m m o n successor in R. T h e clause set generated by t h i s f o r m u l a and Sim2 is i n f i n i t e . However, the special s t r u c t u r e of t r a n s l a t e d f o r m u l a e allows to take the simple u n i t clause

N o t e t h a t serial]ty is already covered by Sim2- For KD there exists m e r e l y a single u n i t clause a n y w a y a n d we are already done. For K D B a f u r t h e r u n i t clause can be derived between S i m 2 and the s y m m e t r y clause w h i c h i s R ( u : x , u ) . T h u s , a single u n i t clause for K D ( n a m e l y R ( u , u : x ) ) a n d t w o u n i t clauses for K D B ( n a m e l y R ( u , u : x ) a n d R ( u : x , u ) ) suffice for the d e s c r i p t i o n of the respective m o d a l logic theory. 4.2 T h e Logics K D 4 a n d S4

as a s u b s t i t u t e for the directedness a x i o m . T h i s can be proved by s h o w i n g t h a t the a p p l i c a t i o n of any other generated u n i t clause can be replaced by several a p p l i cations of . In f a c t , t h i s u n i t clause nicely reflects our i n t e n t i o n w h a t the directedp r o p e r t y is concerned for it states t h a t for any t w o w o r l d s u:x a n d u.y ( w h i c h are b o t h accessible f r o m u) there is a w o r l d u : y : f ( u , u:x, u:y) w h i c h can be accessed by u:x. R is called dense if no w o r l d has a u n i q u e successor in R (different to t h i s very w o r l d ) . By a s i m i l a r a r g u m e n t to t h e above t h i s p r o p e r t y can be s i m p l i f i e d to

Here t h e set generated by A x i o m s and S i m 2 is i n f i n i t e . Nevertheless, its elements have a c o m m o n struct u r e w h i c h is R ( u , u:x 1 . .. : x n ) where n > 0 for S4 and n > 1 for K D 4 . It is thus very easy to f i n d an altern a t i v e clause set w h i c h generates exactly t h e same set, n a m e l y the t w o - l i t e r a l clause - R ( u , v ) V R ( u , v : x ) plus R ( u , u ) for S 4 a n d R ( u , u : x ) for K D 4 . T h e overall effect of t h i s t r a n s l a t i o n is therefore t h a t at least the t r a n s i t i v i t y clause gets s i m p l i f i e d . 4.3 T h e Logics K D 5 , K D 4 5 , and S5

w h i c h states t h a t for any u a n d any u:x accessible f r o m u there is a w o r l d u : g ( u , u : x ) between u and u:x; j u s t as we expected. 4.5 Non-Serial M o d a l Logics

For these logics the generated set consists of the clauses o f the f o r m for and for K D 5 . A g a i n it w o u l d be very easy to find a s u i t a b l e clause set w h i c h generates the same clauses a n d w h i c h is s i m p l e r t h a n t h e o r i g i n a l theory. However, we can do even better if we e x p l o i t a useful result k n o w n f r o m the m o d a l logic l i t e r a t u r e , n a m e l y t h a t w e are allowed t o consider o n l y w o r l d s w h i c h are accessible f r o m t h e i n i t i a l w o r l d b y f i n i t l y m a n y R-steps. T h u s , i f w e i n s t a n t i a t e t h e u f r o m above w i t h i a n d have i n m i n d t h a t any w o r l d whatsoever can be described in t h e f o r m t:y 1 . . . :y m for s u i t a b l e i n s t a n t i a t i o n s of the y i we get R ( u , v) for S5, R ( u , v : x ) for K D 4 5 and R ( u : x , v : y ) for K D 5 . These u n i t clauses f o r S5 a n d K D 4 5 even subsume S i m 2 a n d hence t h e i r respective theories are d e t e r m i n e d by a single u n i t clause. R ( u : x , v : y ) does n o t subsume R ( u , u : x ) , nevertheless it can be used for a s i m p l i f i c a t i o n g e t t i n g R ( i , i : x ) (since a l l other cases are in f a c t subsumed by R ( u : x , v : y ) . N o t e t h a t these s i m p l i f i e d versions nicely reflect Segerberg's discovery ([Segerberg, 1971]) t h a t S5 is characterized by t h e u n i v e r s a l accessibility r e l a t i o n (a non-degenerate cluster; therefore R ( u , v ) for a l l worlds

In non-serial m o d a l logics we are n o t any m o r e allowed to assume t h a t each w o r l d is n o r m a l w i t h respect to the accessibility r e l a t i o n Hence, the s i m p l e u n i t clause f r o m Sim2 ( w h i c h is R ( u , u : x ) ) is n o t v a l i d a n y m o r e since it m i g h t happen t h a t x is not defined on u. Therefore, assume a new u n a r y predicate which is supposed t o denote " n o r m a l i t y " . T h e s i m u l a t o r a x i o m Sim 2 thus has to be changed to Moreover, the i n f o r m a t i o n t h a t certain w o r l d s are n o r m a l has to be reflected i n t h e f o r m u l a m o r p h i s m w h i c h becomes:

U n f o r t u n a t e l y , some of the nice results we got for serial m o d a l logics are lost t h i s way. So, for instance, we get again a n e x p o n e n t i a l g r o w t h i n the n u m b e r o f clauses after t r a n s l a t i o n . Nevertheless, t h e s i m p l i f i c a t i o n s f r o m above s t i l l w o r k p r o v i d e d some N-literals are added at a p p r o p r i a t e places in t h e s i m p l i f i e d clauses. As an e x a m p l e consider the logic K 4 5 w h i c h is characterized by either a single unconnected worlds or a f r a m e a s w e k n o w i t already f r o m K D 4 5 . A n a l o g o u s to the a r g u m e n t for K D 4 5 we get the t h e o r y clause i.e. t h e successors of every n o r m a l w o r l d can be accessed by any w o r l d . A g a i n , the reader interested in m o r e details is referred to [ N o n n e n g a r t , 1992].

Nonnengart

83

4.6

Varying Domains

Recall that we assumed constant domains in the Definitions 2.1 and 2.2. If we want to take varying domains into account a further predicate has to be introduced, namely one which expresses existence of domain elements in given worlds. Thus the relational translation changes to

the clause for K45 and K5, i.e. we have to assure that the FR-variable x is indeed defined on u. This evidently has to be done for any of the above simplifications. We omit details here; the interested reader is referred to [Nonnengart, 1992]. 4.7 4.7.1 Multiple Modalities A Simple Temporal Logic

where the predicate symbol E denotes "existence". The other cases remain as in Definition 2.2. The idea is now to functionally simulate this E-predicate just as we did for theRl-predicate before. Hence, let FE be a functional simulator for E, the formula morphism changes to:

Consider two KD4-modalities and where is responsible for the future and for the past respectively. Moreover assume that there is only one thing said about their connection, namely that the associated accessibility relations (i.e. the "later"- and the "earlierrelation) are converse. Thus we have as a theory axiom: RE(U, V) Rp(v, u) for any two worlds (time instants) u and v. This leads to the following "time-theory":

where means that every x in gets replaced by u:y. Note that we asume that no domain is empty and therefore each world is normal with respect to E. If there are no extra properties given for E then the theory is simply reflected by the simulator axiom E(u,u:x) where the x now ranges over FEUsually, however, one is not really interested in arbitrary varying domains. Often one considers either increasing or decreasing domains, i.e. either nothing gets lost or no elements get newly generated. The axiom for increasing domains is

Because of the equivalence between Rr(u,v) and Rp(v, u), it is not really necessary to consider both predicates, i.e. we can for instance, replace any occurrence of Rp in the translated formula as well as in the theory axioms by RF provided the arguments are reversed. We thus get the theory axioms (where R is taken as a shortform for R F ) :

Note that again the translation from above produces formulae whose clause normal forms do not contain any positive E-literals. We therefore look for a suitable simplification by examining the clauses generated by This results in which states that if x exists in world u then it exists in any world accessible by u as well. Interestingly this theory clause is not only characteristic for increasing domains in the logic KD but also for K T , KD4 and S4. For the logics KD5 and KD45 we can get even simpler axioms because of their simple characteristic frames which guarantee identical domains everywhere but possibly for the actual world. This fact can easily be represented by the unit clause E(u:x, v:y) (where x ranges over FR and y over FE) which states that every domain element (denoted by v:y) is known to any world apart from the initial world i. For i itself we still have the unit clause E ( u , u : x ) which can thus be simplified to E(i,i:x). Decreasing domains can be worked out in a similar manner. We omit here the technical details and just provide the result of the possible simplifications. For the logics K D , K T , KD4 and S4 we get the theory clauses and where y ranges over FE and x belongs to FR and for KD5 and KD45 we get the two simple unit clauses and with the FE element x and the FR element y. In the non-serial case we are again not allowed to ignore the "normality"-property. So, for instance, we get

W i t h this we can now start our simplification process which finally results in:

Whether or not these theory axioms should be prefered to those from above is not merely a matter of taste. It is indeed the case that the search space gets smaller in the simplified case because the positive R-literal of the transitivity axiom can unify with arbitrary negative Rliterals from the translated formula of the theory axioms whereas the latter theory clauses do restrict this. 4.7.2 M u l t i p l e Agents The logic KD45 seems appropriate for the forrnulization of the Belief of Agents since it incorporates consistency of beliefs (seriality axiom) as well as positive and negative introspection (transitivity and euclidity axiom respectively). Now assume a set of agents where each agent's belief obeys the KD45 properties. A first idea would be to add the KD45 theory clause to the clause set for each of these agents. However, since this would mean that the KD45 properties hold over the whole world structure, we would automatically have that also each agent considers his own and the other's agent's beliefs as consistent w i t h positive and negative introspection. If we do

84

Automated Reasoning

n o t w a n t t h i s , w e have t o assure t h a t the K D 4 5 properties o n l y h o l d in the w o r l d s w h i c h are accessible for t h e respective agents. T h i s can be done w i t h t h e help of a new predicate W w h i c h , given an agent a n d a w o r l d tells us w h e t h e r t h i s w o r l d is accessible for the agent or n o t . F o r m a l l y : for a l l agents a and all in

where u ranges over t h e w o r l d s of t h e w h o l e universe a n d x a belongs t o the f u n c t i o n a l s i m u l a t o r o f (agent a's belief r e l a t i o n ) . T h u s the respective a g e n t s t h e o r y o f belief is t h e n given by

if the R-literals in the resolvent can be e l i m i n a t e d by the t h e o r y clauses. We d i d n o t yet i m p l e m e n t t h i s , however, there is some s t r o n g evidence t h a t such a c o n t r o l strategy makes the m i x e d approach behave o n l y s l i g h t l y worse (in the worst case) c o m p a r e d to f u l l f u n c t i o n a l t r a n s l a t i o n . T h i s m e t h o d (as it stands now) can o n l y be applied to m o d a l logics w h i c h have first-order describable accessib i l i t y r e l a t i o n properties. P a r t o f our f u t u r e w o r k w i l l b e t o e x a m i n e m o r e properties and p r o p e r t y c o m b i n a t i o n s w h i c h m i g h t b e useful a n d / o r i n t e r e s t i n g .

References
[ A u f f r a y a n d E n j a l b e r t , 1992] Yves A u f f r a y a n d Patrice Enjalbert. M o d a l theorem proving: A n equat i o n a l v i e w p o i n t . J o u r n a l o f Logic and C o m p u t a t i o n , 2(3):247 - 295, 1992. [Chellas, 1980] B. Chellas. M o d a l Logic: An I n t r o d u c t i o n . C a m b r i d g e U n i v e r s i t y Press, 1980. [Farinas del Cerro a n d H e r z i g , 1988] L u i s Farinas del C e r r o a n d A n d r e a s H e r z i g . Q u a n t i f i e d m o d a l logic a n d u n i f i c a t i o n t h e o r y . R a p p o r t L S I 293, Languages et Systemes l n f o r m a t i q u e , U n i v e r s i t e P a u l Sabatier, Toulouse, 1988. [ F i t t i n g , 1983] M e l v i n F i t t i n g . P r o o f Methods f o r M o d a l and I n t u i t i o n i s t i c Logics. R e i d e l , 1983. [Hughes and Cresswell, 1968] G. Hughes and M. Cressw e l l . A n I n t r o d u c t i o n t o M o d a l Logic. M e n t h u e n , L o n d o n , 1968. [ K r i p k e , 1962] Saul K r i p k e . Semantical considerations on m o d a l logics. In Proceedings of a Colloquium on M o d a l and M a n y Valued Logics, H e l s i n k i , 1962. A c t a P h i l o s o p h i c a Fennica, no 16. [ K r i p k e , 1963] Saul K r i p k e . S e m a n t i c a l analysis of m o d a l logic: n o r m a l p r o p o s i t i o n a l c a l c u l i . Z e i t s c h n f t f u r Mathematische Logic u n d Grundlagen der Mathem a t i k , 9, 1963. [ M o o r e , 1980] R. M oore. Reasoning About Knowledge and A c t i o n . P h D thesis, M I T , C a m b r i d g e , 1980. [ N o n n e n g a r t , 1992] A n d r e a s N o n n e n g a r t . First-order m o d a l logic t h e o r e m p r o v i n g and s t a n d a r d P R O L O G . Technical R e p o r t M P I - I - 9 2 - 2 2 8 , M a x - P l a n c k - I n s t i t u t e for C o m p u t e r Science, Saarbriicken, G e r m a n y , J u l y 1992. [ O h l b a c h , 1989] Hans J i i r g e n O h l b a c h . A Resolution Calculus f o r M o d a l Logics. P h D thesis, U n i v e r s i t y o f K a i s e r s l a u t e r n , G e r m a n y , 1989. [ O h l b a c h , 1991] Hans J i i r g e n O h l b a c h . Semantics-based t r a n s l a t i o n m e t h o d s for m o d a l logics. J o u r n a l o f Logic and C o m p u t a t i o n , l ( 5 ) : 6 9 1 - 7 4 6 , 1991. [Segerberg, 1971] K r i s t e r Segerberg. An essay in classical m o d a l logic. Technical R e p o r t 13, University of Uppsala, Filosofiska Studier, 1971. V o l u m e 1-3. [ W a l l e n , 1987] L i n c o l n A . W a l l e n . M a t r i x p r o o f m e t h o d s for m o d a l logics. In Proceedings of the 10th I J C A I , 1987.

i.e. t h e K D 4 5 a x i o m holds o n l y for those w o r l d s w h i c h are accecssible for agent a. For a l l other worlds we s t i l l have w h i c h states t h a t n o t h i n g else is assumed for the agent's beliefs. F i n a l l y , we have to p r o v i d e the correspondences between W a n d N w h i c h is s i m p l y given b y : (because of s e r i a l i t y ) . N o t e t h a t t h i s technique also allows t o express m u t u a l beliefs of m a n y agents. To t h i s e n d , we i n t r o d u c e a m u t u a l belief accessibility r e l a t i o n which evidently m u s t obey the K D 4 5 properties as well (since each of the agents does). T h u s the s i m p l e u n i t clause (for all agents a) w i l l do for t h i s purpose. For m o r e details on t h i s issue the reader is again referred to [ N o n n e n g a r t , 1992].

Summary

We proposed a t r a n s l a t i o n m e t h o d f r o m m o d a l logic i n t o first-order predicate logic w h i c h allows a considerable s i m p l i f i c a t i o n o f the accessibility r e l a t i o n theory. T h i s approach is some sort of a m i x t u r e between the s t a n d a r d r e l a t i o n a l t r a n s l a t i o n approach and the f u n c t i o n a l t r a n s l a t i o n m e t h o d proposed b y O h l b a c h a n d others. I t shows to behave p a r t i c u l a r l y nice in the a p p l i c a t i o n to serial m o d a l logics. So, e.g., the b a c k g r o u n d theories for the m o d a l logics K D , K D 5 , K D 4 5 , and S 5 get s o s i m ple t h a t even a direct i n c o r p o r a t i o n i n t o the t r a n s l a t i o n m o r p h i s m becomes possible. A c o m p a r i s o n between t h i s approach a n d the f u l l y f u n c t i o n a l t r a n s l a t i o n m i g h t be i n t e r e s t i n g at t h i s stage. For the logic KD b o t h m e t h o d s are i d e n t i c a l as can easily b e checked. For S5, K D 4 5 a n d K D 5 the m i x e d approach e v i d e n t l y works b e t t e r (even if s o m e b o d y is able to find some s u i t a b l e t h e o r y u n i f i c a t i o n a l g o r i t h m for these t h e ories) because of t h e s i m p l e b a c k g r o u n d t h e o r y described by a single u n i t clause respectively. W h a t the logics K D B , K D 4 a n d S4 are concerned, we do n o t have a general answer yet. We have to d i s t i n g u i s h between t h e o r e m provers w h i c h do allow the d e f i n i t i o n of a r b i t r a r y t h e o r y u n i f i c a t i o n a l g o r i t h m s a n d those w h i c h don't. I n t h e l a t ter case a r a t h e r s t r o n g e q u a l i t y h a n d l i n g m e c h a n i s m is necessary for the f u n c t i o n a l t r a n s l a t i o n approach and t h i s m i g h t cause troubles for t h e prover. I n the f o r m e r case the f u n c t i o n a l t r a n s l a t i o n a p p r o a c h w i l l be m o r e efficient in general. Nevertheless, we can t h i n k of some easy c o n t r o l s t r a t e g y for the m i x e d approach i n w h i c h r e s o l u t i o n steps between n o n - i l - l i t e r a l s are allowed o n l y

Nonnengart

85