You are on page 1of 28

VV-OLFGANG

Generalized Net Structures


B ALZEIL ~
AND
of Empirical Theories. II
JOSEPH :D.
~NEED 2

IV. Theories and normal science


In t h e r e m a i n d e r of this p%per we will use t h e concepts described
up to now in order to give a. p a r t i a l answer t'o the following two questions
-partial in t h a t we are concerned only wit,h t h e formal products of science,
n a m e l y theories. These questions are: "~Vha.t is an empirical t h e o r y ? "
a n d " l [ o w do sciences d(.~-elop ? ". B o t h questions are i n t i m a t e l y connect, ed,
indeed, ~he first' question is a v e r y special case of t h e second.
According to K u h n [1] there are two different kinds of periods du::ing
a .~cience's development,: l>(riods <.f n o r n ' a l science a n d periods of scientific
revolutions. B o t h kinds arc characterized essentially b y using the concept
of a ,~cientific c o n m m n i t y , l~oughly, p(riods of norm%l science are those
d u r i n g which a scientific, conmmnfl.y works on a. c o m m o n basis, lm.radigm
or m a t r i x , while .,ci(ntific revolut.ions are characterized b y a. change of
1:his basis, p a r a d i g m or m'ttrix. Ignoring all t'he essentiMly sociological
a n d psychological feat'm'es (,f a scient'ific cmmmmit'y we arc lef~ with
w h a t it produces. This product' is expected to reflect, t,he d i c h o t o m y
normal-revolut'ionary, indeed, if we agree to call t'he product.s o~ a. scien-
t'ific communit,y during' :~ period of n o r m a l science "theories" t h e n t.his
d i c h o t o m y is represented in t h e output's of scientific communities in a
v e r y nat'm'al way.
I n this sect, ion we will describe t,hc products o* scientific comnmnit, ies
dm'ing periods oi~ n o r m a l science a.nd t,hc d e v e l o p m e n t of t'his product's
in such a period. I n S-c. V scient'ific revolutions -- as t h e y arc reflected
in theories - - w i l l be considered.
We do n o t wan~ to r e p e a t what' t'here can be s~dd a b o u t t'hc n%turc
of the c o m m o n basis or pa.radigm which connects scient, ists during nornm.1
science. :For our purposes only one point, is import, ant,. Accepting a. pa.ra-
digm implies s i c w i n g t,hc world t h r o u g h a certain k i n d of glasses a n d
using a terra.in kind oi la.ng~age t'o express "wha.t' is seen". More precisely,
we hold t.hat, ha.ring a pan~digm implies t h e use of a certain set of basic
concepts which are applied to a cert.ain k i n d o~ phenomena.. Thus, an
essential fe'~tm'e of a t h e o r y :~s t h e ont,come of a scientific c o m m u n i t y
working in a f r a m e of one c o m m o n p a r a d i g m is t'hat t h e t,hcory applies

1 This a u t h o r ' s work on this paper was p a r t i a l l y supported by the Deutsche


F o r s c h u n g s g c m c i n s c h a f t ( l ) F G - p r o g n u n K A 407).
2 This a u t h o r ' s work on this paper was carried out during his visit ,~t the Seminar
ftir Philosophic, Logik mtd Wisscnschaftsthcorie der Universitfit Mfi~mhcn nmdc
possible by a grant front i h e Deutsche Forschungsgcmcinschaft.

3 - - Studia Logica, 1//8


168 Wo|fgang Balzer, Joseph D. 8need

a c e r t a i n class of b a s i c c o n c e p t s . T a k i n g a w a y s o m e of t h e s e will m a k e
it i m p o s s i b l e t o use t h e theory, w h i l e a d d u c i n g n e w basic c o n c e p t s will
y i e l d a n e w p a x a d i g m . I n t h e l a n g u a g e of n e t s t h i s m e a n s t h a t a n e c e s s a r y
c o n d i t i o n f o r a t h e o r y - n e t t o r e p r e s e n t a t h e o r y is t h e f o l l o w i n g : All
Mj,'s in t h e n e t h a v e t h e s a m e struc~m'e. T h a t is, T a n d 0 a r e e x c l u d e d
as r e l a t i o n s a m o n g t h e e l e m e n t s of t h e n e t . So t h e o n l y r e a s o n a b l e r e l a t i o n
we a r e l e f t w i t h is a. I n c a s e of a, h o w e v e r , --~ is p r o v a b l y t h e identit;y.
L e t us s t a t e t h i s e x p l i c i t l y .

D20. X is a a-theory-net iff X - - ( N , 4, ~) is a t h e o r y - n e t such


t h a t ~< is a a n d ~ is = o n h r .
F u r t h e r t h i s v i e w of scientific activir ~ e x c l u d e s n e t s t h e basic sets
of w h i c h h a v e m o r e t h a n o n e e l e m e n t , b e c a u s e t h e e x i s t e n c e of s e v e r a l
b a s i c e l e m e n t s w o r k i n g w i t h i d e n t i c a l c o n c e p t s implies t h a t t h e n e t t a l k s
a b o u t d i f f e r e n t k i n d s of p h e n o m e n a a n d in t h i s c a s e w e w o u l d s p e a k o f
s e v e r a l t h e o r i e s r a t h e r t h a n of o n e t h e o r y ; it d o e s n o t e x c l u d e , h o w e v e r ,
t h a t t h e class of p h e n o m e n a in q u e s t i o n b e s i ; n l c ~ ' e d i n t r i n s i c a l l y t h e r e b y
p r e s c r i b i n g h o w t h e b a s i c c o n c e p t s h a v e t o b e a p p l i e d t o it. T h a t is, if
w e r e m e m b e r t h a t "class of p h e n o m e n a d e s c r i b e d " c o r r e s p o n d s t o " r a n g e
of i n t e n d e d a p p l i c a t i o n s " t h e r e m a y b e sub-sets of I s p e c i f i c a l l y c o n n e c t e d
w i t h special c o n c e p t s , i.e. special laws.
W e m a k e t h i s e x p l i c i t in t h e f o l l o w i n g d e f i n i t i o n .

D 2 1 . X is a u~iquely based a-theory-net iff X is a a - t h e o r y - n e t a n d


t h e r e is ",~ ( K , I), ~ .Y s u c h t h a t !8 (N) = ( ( K , I}} 1:~

I n n n i q u e l y b a s e d a - t h e o r y - n e t s all t h e t h e o r y - e l e m e n t s (K,:, I i )
ha~-e t.he s a m e Jll~,p a n d M~, in t h e i r core K i , a n d I i is a l w a y s a s u b - s e t
of t h i s san:e Jllp~,. I f ( K o , I 0 ) is t h e basic set of t h e n e t we m,~y t h i n k
of t h e n e t ' s c l a i m (cf. D18) in t h e f o l l o w i n g w a y . T h e c l a i m e n t a i l s t h a t
Io e A ( K o ) a n d t h a t v a r i o u s n o n - e m p t y , p r o p e r s u b - s e t s of I o ( I i ~ I o ,
I i v~ 0 ) are in t h e A (Ki) p i c k e d o u t b y t h e s p e c i a l i z a t i o n of K 0 a s s i g n e d
t o t h e m (I; ~ A (K~.)). Y~onghly t h i s c l a i m sa.),s s o m e t h i n g a b o u t t h e a r r a y
of n o n - e m p t y sub-sets of t h e 2]I~ c o m m o n t o all t h e o r y - e l e m e n t s in t h e
13 Uniquely based a-theory-nets correspond roughly to correctly applied cores for
theories of mathematical physics in [3] (D37-1, p. 182). The core-nets associated with
thcm correspond to expanded cores for theories of 'mathematical physics (D32, p. 179).
There is however one essential difference which reflects a way in which wc now regard
the treatment in [3] to be fundamentally mistaken. D32) of [3] requires cssentiMly
that special laws be assigned to sub-sets of Mpp (See [5] D9) pp. 130 for a treatment
in the present notation that corrects a technical error of the treatment in [3]). This
means intuitively that we must specify in the conceptual structure of the theory the
"outer boundary" of the range of intended applications for the special laws. In the
present formMism it would correspond to a dcfinitiou of "specialization" (the a-re-
! I
lation) in which .~lpp c .Mpp instead of Mpp = Mpp. This now seems to us not to
correspond to the way special laws are actually employed. We do not, for example,
specify i~ general the sub-set of all kinematicM systems to which the Hooke's law
Generalized net structures of empirical theories. I I 169

n e t . I t s a y s e l e m e n t s of t h e a p p h c a t i o n n e t h r+ a r e a s s i g n e d t o e l e m e n t s
of t h e c o r e - n e t ZT* in a w a y t h a t y i e l d s a t h e o r y - n e t a n d t h a t t h e c l a i m
of e a c h t h e o r y - e l e m e n t in t h e r e s u l t i n g t h e o r y - n e t is t r u e . N o t e t h a t
t h e c l a i m of t h e b a s i c e l e m e n t <Ko, I0> ( t h e " b a s i c c l a i m " ) m a y b e v a c u o u s
( A ( K o ) = Pot(M~v)) and y e t t h e c l a i m of t h e n e t n o n - v a c u o u s b e c a u s e
A (K~) ~ .Pot (Mp~) f o r s p e c i a l i z a t i o n s of t h e b a s i c c o r e a p p e a r i n g in t h e n e t .
T h i s u n d e r s t a n d i n g of t h e c l a i m of u n i q u e l y b a s e d a - n e t s leads u s
t o t h e f o l l o w i n g c o n s i d e r a t i o n s . C o n s i d e r t h e set of all a p p l i c a t i o n - n e t s
c o n s t r u c t a b l e o u t of M~v. T h e s e will b e all p o s e r s w h o s e e l e m e n t s a r e
s u b - s e t s of M v v . :For a g i v e n c o r e - n e t _N*, w h i c h of t h e s e m a y b e c o m b i n e d
w i t h _N* t o y i e l d a t h e o r y - n e t ( n o t n e c e s s a r i l y N ) w h o s e c l a i m is t r u e ?
W e m i g h t r e g a r d t h e set of all s u c h a p p l i c a t i o n - n e t s , 9~(2~)*, as t h e e m p i -
r i c a l c o n t e n t of _N* in a n a l o g y t o t h e w a y we r e g a r d A (K) as t h e e m p i r i c a l
c o n t e n t o f a single core K . W e d e f i n e 2(_N*) p r e c i s e l y as follows.
D22. I f _N is a u n i q u e l y b a s e d a - t h e o r y - n e t w i t h ~ { h r) = {<K, I>}
a n d N* t h e c o r e - n e t a s s o c i a t e d w i t h N t h e n :~ e 9~(N*) iff :~ 9 .Poser (Mpp)
and
f o r all X 9 fXl: X 9 !l~
(ii) t h e r e a r e a: [:~[-+]N*I a n d ~<' _~ (a • 5) s u c h t h a t
( a l ) N ' : = <5, ~<', = > is a a - t h e o r y - n e t , ~ ( N ' ) = {<K, I>}
(a2) (~v')* = iv*
(a3) ( N ' ) + = 13~f
(ad) f o r all X , :Y e [3~f: if X ~<'+ 17 t h e n X ___ :Y
(b) f o r all K e l-N'l: [,.J {XIa(X) = K} c A ( K )
:Note t h a t t h e e l e m e n t s of 9~(_N*) a r e r e q u i r e d b y D22-iia) t o h a v e
a r a t h e r specific k i n d of structl~re as p o s e t s . T h e y m u s t b e h o m o m o r p h i c
t o _N* u n d e r s o m e a. T h e h o m o m o r p h i s m is in a sense t r i v i a l b e c a u s e
we r e q u i r e t h a t we b e a l w a y s a b l e t o " c o l l a p s e " t h e h o m o m o r p h i s m t o
a n i s o m o r p h i s m w h i c h y i e l d s a t h e o r y - n e t all of w h o s e c l a i m s a r e t r u e
(D22-iib). T h i s r e q u i r e m e n t w i t h " t r a n s i t i v i t y of c o n s t r a i n t s " (DO-B3e)
a ssm'es t h a t t h e c l a i m s of t h e e l e m e n t s in t h e h o n m m o r p h i c n e t N ' a r e
also t r u e . ~u c a n n o t m e r e l y r e q u i r e t h e c l a i m s of e l e m e n t s in I ' t o b e
t r u e . T h i s w o u l d n o t a s s u r e t h a t t h e l m i o n of all a p p l i c a t i o n s a s s o c i a t e d
is claimed to apply. Rather wo pick out some sub-set of the intended applications
for Ncwtonian particle mechanics -- I o in our formalisln -- and claim that Hooke's
law holds in these intended applications. The essential difference is that the latter
operation can be (though it must not be) carried out by simply listing what we claim
to be Hooke's law systems without providing an intensional description of them.
The former operation requires that we provide some general, intensional characteri-
zation of Hooke's law systems at the kinematical level. It appears that in this case
Newtonian physicists do not, in fact, do this. There may, of course, be cases where
the range of application for special laws in K~ is described quite generally. But these
may as well be handled in the present formalism by simply describing Ii in the quite
general way. We are indebted to Prof. A. Kamlah and Dr. C. U. Moulines for assi-
stance in clearifying this point.
-170 Wolfgang Balzer, Joseplt D. N~eed

w i t h K e IN*I was in A ( K ) . The constraints m i g h t rnle this out. Intui-


t i v e l y , constra.ints in a core are s u p p o s e d to range over e v e r y t h i n g to
which t h e laws of t h e core a.pply. On t h e o t h e r ha.rid, we need ma.ny-one
a ' s a.s well as one-one a's to adequa:%ely depict t h e relation b e t w e e n 9~(N*)
a n d 9/(N'*) w h e n _NU _N' (T8) If we w a n t 9 / ( N ' * ) _ 9/(N*) we m u s t
a.llow for t h e possibility t h a t applications a.ssigned to distinct cores in N'*
m a y b e assigned to t h e same core of N*. F o r exam.pie, N* might contain
a.n element for distance d e p e n d e n t central forces while N'* c o n t a i n e d
a.s well elements for H o o k e ' s law a n d inverse-square forces. An .~ e 9~(N'*)
w o u l d n o t be in 9~(N*) mfless we could a,ssign b o t h t h e H o o k e ' s law a.p-
])lic,~,tions a.nd t h e inverse-squa.re ~pplic~tions in .~ to t h e dista.nc~ depen-
d e n t central element in N*.
Initially, we h a d required tha.t ~<'+ be identical with t h e inclusion
relation _~ on [.~1- B u t t h e n t h e n e t - s t r u c t u r e of N ' h~4 to reflect t h e
p o s e t - s t r u c m : e of :~ which could be accidental (at least pa.r~ially) in t h e
following sense. If different Xi's e I.~] h a d been d e t e r m i n e d in a w a y
i n d e p e n d e n t front t h e present t h e o r y ' s theoreticM appa.r~tus a.nd if cores
h a d been adjoined to these X:'s b y a t h e n it c o u l d t u r n o u t t h a t some X~
is a sub-set of sonte other X~ w i t h o u t t h e corresponding core Ki being
a, specialization of K j . This might h a p p e n b e c a u s e t h e first step in t h e o r y
construction is ahvays to find some theoretical al)paratus tha.t ca.n deal
with t h e given p h e n o n w n a . Only a.ccidenta.lly one will find o u t sub-set
relations a m o n g cla.sses o~ different phenom.ens which c a n n o t be seen
immediately. So starting with t h e r e q u i r e m e n t (N') + = :~ we arrived at
D22-ii~t4) which sa.sm t h a t t h e a p l d i c a t i o n - n e t associated with N ' is a
s u b - s t r u c t u r e of .~.
Trivial meml)ers of Poset(Mj,~,) axe ruled o u t b y requiring all m e m b e r s
of i.~1 to be n o n - e m p t y . B u t note t h a t t h e a which sq.tisfies D22-ii) for
a, given .~ e Poset(Mm, ) need n o t b e nnique. This m e a n s t h a t .~'s could
g e t into ~I(N*) in several w a y s some of which correspond to " i n t u i t i v e l y
reasonable" wa.ys to a.ssign laws to their m e m b e r s a n d some th,~t " j u s t
h a p p e n to work". F u r t h e r , some could get in o~ly t)y h l t u i t i v e l y ram.c-
eel)table means t h a t " j u s t h a p p e n to w o r k " .
Vfe could t i g h t e n our definition b y providing t h e forma.1 means for
representing t h e "intuitively reasonal)le" w a y s of assigning m e m b e r s of
e Pot"(Mpj,) to m e m b e r s of 15~*1. B u t how we might interpret this
piece of formalism a.part {rom t h e single element of Pot~ t h a t con-
stitutes the " i n t e n d e d a.pplication" of t h e n e t N* is unclear. While we
m a y have an intuitive idea of w h a t is ,~ reasonable a.nd tmreasonal)le
w a y of assigning special l~ws to our " c o n c r e t e " i n t e n d e d applications,
it is less clear t h a t this intuition carries over to all possible intended
,~pplications. Thus it appears reasonabh~, to stick with the looser fornmlation.
H a v i n g defined 9~(N*) we might consider f o r m u l a t i n g t h e "claim"
of .hr in t h e following wa?': ~u e 9i(57"). This is h o w e v e r a weaker view
Generalized net structures of e~pirical theories. II 171

of tlle " c l a i m " of N t h a n t h a t d e f i n e d b y D18). :For t h e c l a i m (as d e f i n e 4


b y D18) of N m i g h t b e i a l s e a n d y e t s o m e o t h e r w a y of a s s i g n i n g t h e
e l e m e n t s of N + t o e l e m e n t s of _N* e x i s t t h a t w o u l d y i e l d ,~ t h e o r y - n e t N '
w h o s e c l a i m (as d e f i n e d b y D18)) was t r u e . Unless we i n t r o d u c e s o m e
k i n d of "ca.nonical ,~ssignment" of special laws t o e l e m e n t s of Pot(Mp~))
in d e f i n i n g 9I(N*) t h i s f o r m u l a t i o n of t h e " c l a i m " of N m u s t a l w a y s
r e m a i n w e a k e r t h a n t h a t d e f i n e d b y D18).
:Even t o s e c u r e t h i s r ~ t h e r loose c o n c e p t i o n of ~he e m p i r i c a l c o n t e n t
of a c o r e - n e t t o p a r a l l e l t h e c o n c e p t of e m p i r i c a l c o n t e n t of a single c o r e
we h a v e t o r e s o r t t o ra.ther c o m p l e x e n t i t i e s . T h a t is, c o r e - n e t s are re-
g a r d e d as s ~ y i n g s o m e t h i n g a b o u t p o s e r s o v e r Mm,-singling o u t ~ sub-se~
of t'~set(Mp~,). Is t h e r e n o t fl, wa.y we ca.n b r i n g t h e c o n c e p t of t h e e m p i r i c a l
c o n t e n t of ~ c o r e - n e t closer t o t h e c o n c e p t of t h e e m p i r i c a l c o n t e n t o f
a si~gle core a n d r e g a r d t h e m b o t h as s a y i n g s o m e t h i n g a b o u t s u b - s e t s
of Mj, v -sing'ling o u t a s u b - s e t of Pot(M~)~))? One w a y t o n p p r o a c h t h i s
is t.he following. F o r t h e c o r e - n e t 2V*, let A ( _ Y * ) ~ Pot(M~,p) b e d e f i n e d
as follows : X _ Mpj, is in A (N*) iff sub-sets of X c a n b e u s e d t o c o n s t r u c t
a n l e m l ) e r of Poset(Ml,1, ) w h i c h is in 9I(N*). l ~ o u g h l y , X e A(N*) iff t h e r e
is s o m e w a y t o assign n o n - e m p t y s u b - s e t s of X to t h e cores in N* to p r o d u c e
n t h e o r y - n e t w h o s e c l a i m is t r u e . W e m a y d e f i n e A (N*) p r e c i s e l y as follows.
D 2 3 . If N is a u n i q u e l y b a s e d a - t h e o r y - n e t a n d N* t h e c o r e - n e t
as.~oeiatcd w i t h N t h e n A (N*) =,~{X e Pot(Mm,)[ there is a I.~l e Pot2(X)
s~wh that
(i) X e I.~l
(ii) ~or all Y ~ [.~1: Y E 9.1~
(iii) (,IXl, - - - ) ~ 9~(N*)}
I f <Ko, Io~> is t h e basic set of a l m i q u e l y b a s e d a - t h e o r y - n e t N, t h e n
ave m i g h t c o n s i d e r f o r m u l a t i n g t h e " c l a i m " of N as: Io e A(N*). T h i s
vi,,w of t h e " c l a i m " of N is still w e a k e r t h a n c l a i m i n g t h a t N + e ~ ( N * ) .
F o r , e~-enthough N + ~ 9~(N*), t h e r e m i g h t b e s o m e o t h e r w a y of ca.rving Io
nl) i n t o a p o s e t o v e r M~w w h i c h c o u l d b e a s s i g n e d t o e l e m e n t s of N* in
a w a y t l m t .~ecured its m e m b e r s h i p in 9/(N*). T h u s t h e f o r m u l a t i o n of
t l w " c l u i m " of _hr c o n s i d e r e d h e r e is " d o u b l y " w e a k e r t h a n th:~t d e f i n e d
b y D18)'~.
la A (3T*) corrcsponds roughly to ~V~ in [3] (D35, p. 181) and Ae(E)in [5] (D10b,
p. 133). The difference is that A (N*) can not be employed in the same way to cxprcss
empirical claims for specific ranges of intended applications. To claim that I e _u
is to claim something stronger than to claim that I 9 A (N*). Roughly, the dcfinition
of the expanded core for a theory of mathematical physics E has packed iato it a cano-
~dcal assigumcnt of non-theoretical structurcs to special laws which removes the
ambiguity about how elements of I are to be assigned to special laws. We have con-
sciously elected not to introduce this (see Note 13) on intuitive grouuds. Though
tlw treatment of [3] appears to provide ~ "tighter" conception of the empirical conteut
of ~ ct, rc-uct N* (or t.he ,~nalogous expauded core E) we aow regard this "tightuess"
a.~ a spurious and n)isi:,~ding attraction h~ving no sound i~xt~uitive basis.
172 Wolfgang Balzer, Joseph D. Sneed

The following t h e o r e m notes t h a t b o t h concepts of t h e "empirical


c o n t e n t " of a a-theory-core-net have t h e p r o p e r t y t h a t e x p a n d i n g t h e
net narrows down t h e empirical c o n t e n t of t h e net.
TS. I f 1~ and 1~' are uniquely based (~-theory-elements such that ~ (N)
= !D(s and 1~ E lq' then
(a) 9/(_~'*) ~ ~/(2r
(b) A ( h r'*) ~_ A ( ~ * )
Uniquely based a-theory-nets are n o w used to define t h e concept of
a K-theory. K-theories are those formal structures which m a y be used
to describe w h a t is produced by a scientific c o m m u n i t y during periods
of n o r m a l science. We call t h e m K-theories to indicate t h a t we ~mder-
s t a n d K u h n to be describing entities of this sort in (1).
D24. X is a K-theory iff there exist Ko, I v, N v a n d ~V such t h a t
(1) X = <Ko, ~v, 2V~, ~ >
(2) (Ko, Iv> is a theory-element
(3) N v is a a-theory-net such t h a t !B(N~) = {(Ko, I~>}
{ ,(a) Nisa a-theory-net }
(4) ~ = h r (b) there is an lo, Iv ~_ I o s u c h t h a t ~ ( ] r {<Ko,Io>}
(c) 2G_~N
A K-theory consists of a basic conceptual struetuxe K0, a set of paxa-
d i g m applications of this basic conceptual s ~ u c h t r e Ip, a specification
of how this basic concept~lal structure is applied to t h e p a r a d i g m set
of applications -- the p a r a d i g m theory-net hr~ -- a n d a set of "admissible"
expansions of this p a r a d i g m net, W.
We conceive of Ip as t h e actual initial set of successflfl applications
of t h e conceptual a p p a r a t u s Ko. F a r t h e r "potential" applications en-
visioned by t h e "founders" of t h e scientific tradition associated with I/7o
axe excluded. Likewise elements of ~4r are conceived as actual applications
of Ko to some range of intended applications I0 which have been explicitly
"examined" by t h e scientific c o m m u n i t y . E a c h could represent t h e state
of development of t h e scientific ~ a d i t i o n at some particular time sub-
sequent to its founding. The full range of d e v e l o p m e n t potentials is in-
cluded in JV. W h e t h e r there is some single _T0 at t h e t o p of all members
of JV which includes all others a n d represents the range of intended ap-
plications for Ko is a question we leave open. xs
We require t h a t hrp be a sub-set of all nets in ~ b u t n o t necessaxily

x6 The existence of a single i o - - t h e range of intended applications for K o --


was stipulated in earlier treatments of "theories in the sense of Kuhn" ([3], p. 293;
[4], p. 222). However Stegmttllers treatment of "autodcterminism" of the range of
intended applications in [6] suggests that other possible d e v e l o p m e n t paths should
at least be countenanced.
Generalized net structures of empirical theories. I I 173

~n i n i t i a l p a x t . T y p i c a l l y , hTp will n o t b e a,n i n i t i a l p ~ r t b e c a u s e t h e ex-


p a n s i o n of _Np will conta.in a,d d i t i o n M ~ p p l i c ~ t i o n s w h i c h will b e c o l l e c t e d
i n t o 10 a,t t h e t o p of t h e e x p a n s i o n . W e c o n c e i v e of single e l e m e n t s of ~4"
~s r e p r e s e n t i n g ~ c t u u l a p p l i c a t i o n s of K o . T h e p o t e n t i a l r ~ n g e of s u c h
9~ c t u u l ~ p p l i c u t i o n s is t h e full s e t JV. T h u s t h e I o a t t h e t o p of e ~ c h e l e m e n t
in .Zz r e p r e s e n t s o n l y t h e c o l l e c t i o n of ,~ll t h e " u p - t o - n o w " s u c c e s s ~ f l
~ p p l i c ~ t i o n s of K o - - n o t t h e f~fll r ~ n g e of i n t e n d e d or h o p e d - f o r a,ppli-
c a t i o n s . T h e la,t t e r is c ~ p t u r e d b y t h e full ra,nge of I o ' s a p p e a r i n g in m e m -
b e r s of ~V. ~6 N p " a . n c h o r s " ~V b u t n o t n e c e s s a x i l y b y f i x i n g t h e t o p of
9~ll its e l e m e n t s .
M o r e e x p l i c i t l y , w e ma.y t h i n k of t h e a~lmM h i s t o r i c ~ l d e v e l o p m e n t
of ~ K - t h e o r y a,s :~ s e q u e n c e :

~Vp, ~u N2, . . . , iV9

w i t h N t e JV a.nd N~ _E N t. E ~ c h 2q t r e p r e s e n t s t h e s t a t e of d e v e l o p m e n t
o f t h e K - t h e o r y ~t t i m e t - - w h ~ t t h e t h e o r y ' s p r a c t i t i o n e r s a,r e prep,~red
~ t t t o a,s s e r t as a.n empirica.1 c l a . i m ) :
T h e historica.1 d e v e l o p m e n t of t h e K - t h e o r y is cumula, t i v e a.t (t, t + 1 )
iff

i .e. ~u is a. p r o p e r e x p a . n s i o n of N t. Thin m a y occm" in o n e or t h e o t h e r


o f owe w a T s , o r b o t h wa.ys s i n m l t a . n e o u s l y . :First, t h e torn,1 r a n g e of ~p-
plic,~tions is s i m p l y enla,rgcd w i t h o n r p o s t u l a t i n g a n y n e w special l~ws.
This con'esponds to
(5~) * = (~V+~) *

b u t (K~, I ~ ) a ( K i , 1 ' + " i.e. l i I t+~ f o r s o m e K~ e IN*I*. T h a t is, s o m e


pla, ce in ~he n e t N t n e w a,pplica.tions w e r e s i m p l y ~ d d e d t o y i e l d ,~ n e w

le The treatment here lodges all the "potentiality" ill thc concept of a "theory
in the sense of Kuhn" in the set ~ ' . Each clement N e . V is a net representing an
array of concrete apphcations of the theory's conceptual apparatus and specializations
of it to a well-specified, empirically investigated range of applications. In earlier
treatments "potentiality" was also lodged in the l ' s associated with expanded cores.
No real problems are solved by this change since the same problems ia describing
I ' s conceived as including potcntial applications (See [3], p. 253) now appear in des-
cribing ~P. The only advantage is that the difficulties i~ describing "paradigm de-
termined resemblance" for applications and ways of dealing with applications axe
formally lumped together.
Iv We neglect here the possibility that different members of a scientific community
built around t h e / / - t h e o r y may not completely agree about the statc of the theory's
development at t. To handle this we might let there be a sequence of Nt's for each
member of the community and then consider the comlnunity's history to be the
intersection of the individual sequence. This of course ignores the fact that in the
politics of scientific communities (as in others) some members' views have more weight
than others.
174 ll'olJ'.qa.ng Balzer, Joseph D. S'~eed

th(,ory-elenmnt' in N t+~. Not'e t,hat, N + #- N ~+t because ~\'~+~ has t,he core K~
a.1)pea.ring ~wice - - o n c e w i t h I~ a n d once wit, h II. +t - - w h i l e it ~ p p e a r s
o n l y once in ~Y+. The a d d e d ~pplicas will alwa.ys 1)e eollect'ed at, t,he
401) of t h e n e t so thn.t, II, c. .,1~+~. Second, t'he t,ot'al ra,nge of a.pplica.t'ions
nla.y tools.in u n c h a n g e d , i.e.

b u t addit,ional special hlws are 1)ostIflat,ed to (lcal w i t h t,hem, i.e.


(N~) * c ( X ~ ') *.
][ere we lmve (-ffe(.tively said somet'hing more-son~ething st,ronger in
1:15(, sense of T8) --al)out, t h e s a m e r a n g e of al)plieat'ions. Fina.lly, 1)ot'h
t'h(,se m o d e s of (.un).ulat,iYo develol)m(mt, m a y occur s i m u l t ' a n e o u s l y at,
(,t, t - - 1 .
It, sh(mld be no*(,d Chat', while e n n m l a t i v ( , develol)ment, for K - t h e o r i e s
m a y be nnaml)ioalm~sly described, t,here is no Izamrant,ee t'ha,t, *he a.etnm.l
de~'elolm).ent of K - f h e o r i e s o~'(,r t . i m e - n o r m a l s c i e n c e - is, in fact,,
(nmmlat'iv(,. N o r m a l science (.onld de~-elop in "fit,s a n d st,art.s" o n l y ex-
hil)itin~ eunmh',ti~-e f(,atnr(,s, if at. all, in t'he long rim.

V. Reduction and scientific revolution

lfa.vin~o' descril)(,d h o w t,he 1)r~dncts of a scient,ific e o m n n m i t . y r(.flect,


a ([e~'elopmen~ dm'in~' periods o~ norm.aJs seience t,here r c m ~ i n s t,he t,ask
of doing so wifll lmriods of s(.i(,nt,ific r e v o l u t i o n s Ir :r scient, ific
re~'olu~ion eom.e,s n I) wit'h a. n e w p a r a d i g n t which promises success. Aga.in,
w(, l(,a~c aside all soeioh)gieal a n d l)syehological fea.Mu'es. T h e n e w l ) a r a d i ~ m ,
t h e n , ix exl)r(,ssed ILY lleW lmsic concept,s, i.e. b y ~ n e w k i n d of p.b~sscs
Chrough which tls(, wm'ld ix seen. Quite n:~talra.lly t h i s c h a n g e of t h e b,~sic
con(.epCs will 1)(' connect,ed wit,h n e w laws. So t h e reslflt of ~ scienl;ific
r(,~-olu~ion as reflcct(,d in (,ml)irica.l theories will l)e a. t'ra.nsit,ion fTOm
a K-theory T <K0, Iv, N v, :~V. t,o ~ I(-theorv T' ",;.K0,I0, 5~,, .

~ o m e ha%-c arIz'u('d t'h~t, sueh t'ra.nsit,ions ~dtiat'(, t h e ide:~ of "(mll)il'ic~l


seien('e" ns a enntnla.~iv(', l)rog]'essive elitert)rise a n d ])erh~ps oN-ell }5.6
a rat'ionat (,nt,erprise. T h e y a.roate t.hat, since t,he t,heories in quest,ion a.re
n o t coml)a.rable on(, ea.nnot, m a k e intellig'il)le judgentent,s a.1)out, t h e i r
relat,ix-e cffee~i~-eness m~d t,hus t,he a d o p t i o n of "~ n e w p a r a d i g m "rod
t,he K-t,heory assoeial, ed wit,h i~ lugs a st,rong t,oueh of irr~t,ionMit,y a,l)out
it a k i n 1)t-rha.1)s to a. relio'ious conversion. St, eo'miiller, howe~,er, ha.s a.rg'ued
([511 Kal). I X a n d [6]) t,ha.t, one(, t'he full e o m l ) l e x i t y of K-t,heories ha.s
l)eeome ~pl)arent, ~h(, redl~(.$ion rcla.t, ion ntioh~ lm ~he a.t)propri~e f o r m a l tool
E(n' d(,l)iet,in~' t,he relation be~we('n t w o K-t,heori( s sel)qra.~ed I)y ~ scientific
r(wohtt'ion in whi(,h we int, u i t h ' e l y c l a i m t h a t "l)rogress '' lm.d l).ven n m d e .
][ere we deseril)e 1-15(, l m r e l y f o r m a l possil)ili~ies for e x t e n d i n g t h e re-
du(.tion r e l a t i o n 1)('tw(,en theorT-elcn~.ents t,o "/foist'ion b e t w e e n K - t h e o r i e s .
Generalized ,tet structures of empirical theories. II 175

The m o s t general relation here is t h a t of one t h e o r y - e l e m e n t ' s re-


ducing to a n o t h e r . I t m a y hold b e t w e e n t h e o r y - e l e m e n t s with c o m p l e t e l y
lmrela.ted conceptua.1 ,%pparatus (different theory-element-cores). I t m a y
be specialized in interesting w a y s and m a y be used to define other in-
teresting relations a.mong t,heory-elenwnts. Thus, t h e Q-relation defined
a m o n g t h e o r y - e l e m e n t s in Sec. I I suggests itself ~s t h e one to be e x p a n d e d
to a - t h e o r y - n e t s in an obvious way. ~Vc have only to require t h a t a 0-re-
lation holding b e t w e e n t w o basic t h e o r y - e l e m e n t s be preserved u n d e r
the speci~fiizations of the nets built on these basic elements. L e t us consider
these " p r e s e r v a t i o n " properties.
One m i g h t suspect t h a t , once a rcdnction relation had been discovered
l)etwe(n t h e o r y - e h n t e n t s , it. conld be preserved in t h e following s(~nse.
If R rednces T' t.o T ~md Tt is .~ome si)eciaJization of T' then t,here will
ahvays be some st)ecia.liza.tion T, of T such t h a t R reduces T~ to T2. Ro-
nghly, it should ahvays be possible to replicate the effec~ of a n y special
laws in t h e r e d n c e d t,h(,ory b y corresponding sp(cial laws in t h e reducing
t,h eory.
F o r b o t h r e d u c t i o n "rod strong r e d n c t i o n this is so, prou one is
n o t at, all fi~stidions a b o u t w h a t connts as a. special la.w of t h e reducing
t h e o r y . These facts are s u m m a r i z e d in the following theorem.

T9. I f T' a,n,d T are th.eory-eleme,~.ts and RD(R, T', T) (RD(R, T', T)
the.n, for all T~aT' there is some T.,aT s,~tch that RD(R, T~,/L.) (RD(R,
T1, T~)).
These facts a b o n t preservation of reduction relations t h r o u g h speci~fli-
zation sue'test, t h a t for t h e o r y - n e t s N ' and N the forma.l a.spect of re-
duction is COml)letely c~ptured 1)y a reduction b e t w e e n %he b~sie elements
T' :~nd T. B u t this ignores the intuitive rc'qnirement tha.t we cadmot be
couq)letely undiserindnatin~ in choosing specializ~tions of T which "re-
d u c e " given specializations of T'. RoughlT, we cannot use arbitrarily
c o n t r i v e d special laws of T to tel)reduce the effect of special h~ws in T ' ;
w(' must, use special laws of T t,ha.t a.lready h~ve some "sta.nding" within
t h e l a ~ e r theory. F o r ex{mtple, if we simply required t,heoretica.I speci~fii-
zations t,hen t,he preserva.tion of reducibility under R through speciafii-
zati(m ca.n no longer be esta.blished.
T10. There exist theory-elements T' and T such that RD(R, T', T)
(RD(R, T', T)) a,n,d for some T1crthT' there is q~o T~at,~T su,eh th,at R D ( R ,
T,, T.,_) ( R D ( R , T1, T,,_)).
A n o t h e r w~y to handle t h e intuitive r e q u i r e m e n t t h a t special laws used
in establishing redncibilib" b e t w e e n K-theories h a v e some standing in
t h e t,heory is simply to regaxd r e d u c t i o n as a relation b e t w e e n specific
theory-nets, lr t h e ~-rel,~tion n m s t reduce t h e basic element T'
176 Wolfgang Balzer, Joseph D. Sneed

t o t h e basic element T a n d e v e r y t h e o r y - e l e m e n t of hr' to some t h e o r y -


-element of h r . The following definition m a k e s this precise.

D25. If hr' a n d hr axe a - t h e o r y - n e t s a n d T' a n d T axe t h e o r y - e l e m e n t s


such t h a t !D(hr') = {T'} a n d !B(N) -----{T} t h e n R (strongly) reduoes hr' to
~v (g~(R, iv', hr), (g~(R, iv', hr))
iff

(1) I~D(R, T', T) (RD(R, T', T))


(2) for all T~ e Ihr'l t h e r e exists some T~ 9 Ihrl such t h a t RD(R, T~, T2)
(RD(R, T,, 2~)).

R e d u c t i o n relations b e t w e e n t h e o r y - n e t s m u s t satisfy some r a t h e r


strong requirements. R o u g h l y , the same ~-relation m u s t serve to r e d u c e
e v e r y m e m b e r of one n e t to some m e m b e r of t h e o t h e r net. No a r b i t r a r i l y
c o n t r i v e d t h e o r y - e l e m e n t s are p e r m i t t e d . Clearly, such relations will
n o t always exist. E v e n if we can a l w a y s r e d u c e t h e basic elements of
t h e nets (exept in highly c o n t r i v e d examples) t h e r e is no g u a r a n t e e t h a t
t h e same relation t h a t reduces t h e basic elements will serve to r e d u c e
a specialization of one to some m e m b e r of ,~ fixed set of specialization
of t h e other.
B u t t h e r e is reason to t h i n k t h a t r e d u c t i o n b e t w e e n specific t h e o r y -
-nets is simply n o t t h e relation we are interested in. l ~ a n y t h e o r y - n e t s
(in t h e course of time) m a y be associated with t h e same K - t h e o r y . Intui-
tively, we t h i n k t h a t r e d u c t i o n is a t i m e - i n d e p e n d e n t relation t h a t should
hold b e t w e e n t h e theories themselves. An obvious m o v e is requiring
t h a t a formal r e d u c t i o n relation b e t w e e n theories r e d u c e (in t h e a b o v e
sense) all pairs of t h e o r y - n e t s for t h e theories. B u t this is t o o strong.
V e r y "big" t h e o r y - n e t s of T' will n o t generally r e d u c e to v e r y " s m a l l "
t h e o r y - n e t s of T. A more satisfactory m o v e w o u l d require a formal re-
d u c t i o n relation only b e t w e e n t h e t h e o r y - e l e m e n t s associated w i t h t h e
p a r a d i g m theory-nets. These nets r e m a i n fixed t h r o u g h t h e t h e o r y ' s
d e v e l o p m e n t and, p r o v i d e d t h e nets themselves are non-trivial, t h e exis-
t e n c e of r e d u c t i o n relations will b e non-trivial. W e will call this k i n d
of reduction " p a r a d i g m - r e d u c t i o n " (see D26) below).
I t should be n o t e d t h a t a r e d u c t i o n relation b e t w e e n p a r a d i g m t h e o r y -
-nets can always be "trivially" e x t e n d e d to expansions of t h e s e nets.
I f we t h i n k of hr' being " e x p a n d e d d o w n w a r d " b y specializing some of
its m e m b e r s , t h e n we can always find some "downwaa.'d e x p a n s i o n "
of h r such t h a t 0 reduces hr' to h r. T h e o r e m 9 assures this, p r o v i d e d we
a d m i t a r b i t r a r i l y c o n t r i v e d expansions into t h e " d o w n w a r d e x p a n s i o n "
of hr. If we t h i n k of t h e p a r a d i g m t h e o r y - n e t s as establishing (among
o t h e r things) t h e " f o r m s " of special laws p e r m i t t e d in applications of
Generalized net st~'uctures of empi~,ical theories. II 177

~he theory, this arbitraxiness is not objectionable. Objectionable '%d hoe"


special l~ws h~ve already been "filtered o u t " by the paradigm theory-net.
This is made precise in D26-b).

D26. If T' and T axe K-theories then


(u) R (strongly) paradigm-reduces T' to T (t)-- ~ ( R , T', T),
(P -- 9 ~ (R, T', T))
iff

(b) R (sC-rongly) complete reduces T' to T


iff
(1) P - - ~ ) ( R , T', T) ( P - - ~ ) ( R , T', T))
(2) for all N' e ~ ' there is some N e~4r such t h a t ~ ( R , N',N)
(~Z)(R, ~V', ~ ) ) .

T l l shows ~hat this kind of "complete" reduction is not stronger


t h a n paxadigm reduction for a-theory-nets, but is stronger for ath-theory-
-nets. This follows immediately from T9 and TI0.

T l l . (a) For all R e f~B, if T and T' are K-theories and P - - ~ D i R ,


T', T) then R strongly completely reduces T' to T
(b) T l l - a ) becomes false when a is replaced by ath.
W e conclude with a remaxk on relations among K-theories which can
be stated in a natural way just using the concept of nets. We might call
nets consisting of K-theories "second-order nets".

T12. (a)
I f S e ~IR is a .finite set of K-theories and
~< = {{x, y} ~ S x S I 3 R ( P - - ~ 5 ) ( R , x, y))}
4 ' = {{x, y} e S x SlUR(F- ~ ( R , x, y))}
= { { x , y } e s x S Ix<~y and y ~ x }
,-~'---- { { x , y } e S x S I x ~ ' y and y ~ ' x }
then {S, 4 , ~-~} a~d {S, ~', ~-~'} are nets.
(b) Let S = (T1, ..., T~} be a set of K-theories and {T~} ----~ ( N ~ )
for i ~ n. I f {{TiIi ~ n}, (~, ----} is a a-theory-net and
{ { T i , TS> e .,s' x s I.T,: a-T,.}
< :
,--, = { { T , ,
T p e S x S I T~,~T i and T i aTi}
then {S, 4 , ~") is a net.

T12-a) says t h a t K-theories form nets under the "reducibility"-


-relation while T12-b) says that, whenever the basic theory-elements of
the paradigm nets of K-theories comprise a a-theory-net, the K-theories
themselves have a net structure.
178 Wolfga~g Balzer, Joseph D. Sneed

Mathematical appendix
I t m,~y be helpful to sketch tile proofs of the theorems fornml,~ted
in the text, pa.rtly because t h e y arc n o t completely trivial a n d p a r t l y
beca.use this leads to ~ b e t t e r u n d e r s t a n d i n g of the set-theoretic ~pp~ratus~
we developed.
TO. For all M, X , R, R e 9Jl:
(a.) i.f R ~ ( M "< 5r) and R: M ~ N then _R: P o t ( M ) ~ P o t ( N )
(b) if rd(R, M, N) then R: Dzz(R)->Pot(M ).
P1moF: (a) Cleaxly, D s ( R ) = P o t ( M ) a.nd D H ( R ) ~_ Pot(N). If
: X , Y . e / ~ t h e n b y D0-B4) Y = {y !=Ix e X(y = R(x))}. So Y is n n i q u e l y
d(.terntined b y X.
M Z

(b) B y definition R ~_ D H ( R ) ",,<Pot(M) a.-nd D I ( R ) =-: DH(]~), DH(]~ ),


c_ l'ot(M). W e show t h a t ,Z,r X e 2~ and ~Z, ]", e ~ intl)lv X = Y.
F r o n t D0-]~2d) and D0-B6) we obtain
(1) (,z, x , e _~ ~ (x, z E k ~-~ 3 ( . ( x ~ , z A V . , , ~ x(::.,., , , ( . , , ) E ,r
9 v"

(2) <Z,Y eR~-+.(Y,Z e]r eR))


(2) implies t h a t
(3) Vv e Z 3 y e Y(c'(y) v).
N o w let x e X . Th(,n, b y (1)
(4) (x, c(x)i e R a~bd c(x) ~ Z. YVe a.pl)lS" (3) with e(x) instead of v
a.nd ol)ta.in

(~) 3.,/~ Y(c'(:,n = c(,)).


As 4~/, C(y)', e R, we ha.~,~e showed tha,t with (4) a n d (5): .~'. , c ( x ) e Ir
, \
A ( y , c ' ( y ) , e R A c ( x ) = c'(y). B n t this iml)lies a, = y, h(.nce x e Y and
X c Y 1)y D0-B5c). To l)ro~e t h e con~:erse let. y e Y. As a.t)ove : {y, c (3) ' e R ,
-- t 7

c'(y) e Z and Vv e Z 3 x e X ( c ( x ) = v). If we substitmt, e c'(y) for v in t h e


last :[ornmla. we obtain =ix e X(v'(y) = c(x)). So (y, #(Y)I e R A ( , , ' , c(x)"
e/r = C(X) which, again b y D0-B5e), implies .r = y. Thus y e X
and Y ~ _ X .
T1. If T and T' are th,eory-eleme.~ds a~d T ' a T th.en
(a,) I' _~ I
(b) A (K') ~_ A (K)
(c) if I E A (K) th,e~, 1' e A (K')
PlmOF: (a.) is a trivi~d consequence of D7-6).
(b) Let X e A ( K ' ) . Then there is a.n Y' ,~nch tha.t ~ ' ( Y ' ) = X and
Y' e P o t ( M ' ) n C ' . F r o m T ' a T we obtain M ' _ M ~md C' _ C, so P o t ( M ' ) n
Ge,neralized net s t r u t t u r e s oJ" em, p i r i c a l theories. I I 179

nC' ~_Pot(M)nC. Thus we h~ve f o l m d a.n Y e P o t ( M ) n C such t,ha~


~(Y) = X, t h a t is, X c A ( K ) .
(c) ~Ve show t h a t VX, Y ( X e A ( K ) A Y _ ~ X A Y e O . I ~ - + Y e A ( K ) ) . The
desired result will t h e n follow f r o m I ' _ I. Let, X, :Y be given wit,h X e A (K),
Y _~ X a.nd Y e ~t. Then t h e r e is a Z such t,ha.t + (Z) = X a.nd Z e Pot (M) n
r~C. Let, U: = {x e Z I r ( x ) e Y}. Then U ~_ Z a . n d + ( U ) = Y. F r o m D0-B3c)
we get U ePot(M)(~C a n d t,hus Y e A ( K ) .
T2. I f I" a++d T are theory-eleme.~ts a++d RD(R, T', T) then /~D(i~,
T', T)
PROOF: D10-1) nnd D10-4) give us 1) a.nd 3) of D9). I t rcm,~ins to
' 3 '
show D9-2). L e t <X, X ) ePot(:lvv) xPot(Mj, v) such t h a t X 9
X ' e g J ~ a n d q"X~'
,X) e ~. Suppose we could consVruct Y' a n d Y such theft
(1) ~'(Y') = X ' , Y' 9 9J~, Y ePot(M)nC a,,~d <Y', I% eT~.
Then by D10-2) we could infer Y' ePot (M') n C ' A V(Y') = X', i.e. X' eA(K').
So a.ll we ha.re to do is to find Y' a n d Y satisfying (1).
X c A ( K ) iml)lies t h a t there is a,n Y~ suelt t}la,t ~ ( Y x ) = X a.nd lrx
ePot(M)r~C. So for each x e X there is at, lea.st one y e Y~ w i t h r(y) = x.
:Now with t h e a.xiom of choice we find a sub-set Y of Y: such t,ha.t for
,~ll x e X ther(; is ('xa.ctly one y e Y with r (y) -- x. I n ot,her words :
(2) r: I ' ~ X .
Y ~_ 1"~ e l)ot(M)nC iml)lies
(3) Y ePot(M)(~C because C is tra.nsit,ive (DO-B3c).
<X', X e/~ yi('lds ~ f u n c t i o n c: X ' ~ X such t h a t Vx' e X'(<x', c(x')> ~ R)
thus

lVe define I"': = ]~(Y) : = {Y'I~Y e Y(<y', y> eR)}. Let, y e Y. We hnve
y e M b(ca.use of (3), r(y) ~ X by (2) a.nd <5(r(y)), r(y)> e k b y (4). W i t h
x ' : = g(r(y)) a n d x : = r(y) this becomes
(5) <x', x> e 1~, y e 3/ a~d x = r(y),
so f r o m D10-3) we o b t a i n an y ' e M ' such tlu~t
(6) <y', y> e R and r ' ( y ' ) = x'.
XeA(K) is n o t e m p t y , so Y e 9 3 l a.nd, b y (5),
(7) 1~' ~ 9)~. N e x t we prove
(s) e --

Let, y be given. I n t h e proof of (6) c(r(y)) wa.s called x' :rod b y (6): r'(y')
= x' = ~(r(y)). Aga.in b y (6) <y', y> e R, so y' = / ~ ( y ) because of D10-1)
a n d D0-B50). Thus we h:wc ~"(_~(y)) = g(r(y)) which is (8).
(9) i:'(Y') =- X'.
180 Wolfga~4] Balzer, Joseph D. 8need

Let x' e X l
Then, using (2) a n d (4) t h e r e is y e Y such t h a t x' = ~(r(y)).
:But ~(r(y))= r' (/~(y)) b y (8) a n d R ( y ) e 17' b y definition of :Y'. So
x' e~'(~Y'). To see t h e converse. L e t z e V(17'), i.e. z = r'(y') with y' ~ I r'.
B y definition of Y' t h e r e is an y 9 :Y with y' ----/~ (y). So z = r'(/~ (y))
= c(r(y)) b y (8), a n d (2), (4) yield ~(r(y)) e X'. H e n c e z e X'. I t remains
to show t h a t <Y', ~ Y ) e / L W e claim t h a t
(10) /~*: 17.-.Iz' where _~* is _~, restricted to Y.

/~* is a flmction because of D10-1) a n d D0-Bhc). The proof of (6) shows


t h a t /~* is defined for e v e r y y e Y. B y definition of 17', ~* is surjective,
i.e. Y' ----D n ( R * ). W e p r o v e t h a t y ~ Yl, Y, Yl 9 :Y implies
(11) .~(r(y)) :/: c(r(yl)). Suppose, R*(y) ---- _~*(yl). T h e n U(R*(y))
= e'(-~ (Yl)) which t o g e t h e r w i t h (8) yields
(12) ~r(y)) ---- ~(r(yl)) in c o n t r a d i c t i o n to ( l l ) .
This proves (10). Now let C: = ~ * . T h e n e: Iz'~-~17 b y (10) a n d Vy' e 17'
(<y', v(y')} e R).
T3. The relatio~s Q, a and % are independent.
PI~OOF: F o r e~ch pair <x, y}, x, y e {Q, a, %}, x r y we h a v e to give
examples of t h e o r y - e l e m e n t s T, T' such t h a t T ' x T a n d -~(T'yT) and
7 (TyT').
l.) <x, y} = <~, ~}
L e t a, b e ~ snch t h a t a r b a n d T: = <<{<a, a}}, {<a, a}}, Id, {<a, a}},
{{<a, a}}}>, {<a,a}}> and T':----<<{<b}}, {<b}},Id, {<b}},{{<b}}}>,{<b}}>.
It is easily checked that T a n d T' are t h e o r y - e l e m e n t s a n d n e i t h e r T' a T
nor TaT' because {<a, a}} $ {<b}} a n d (<b}} $ {<a, a}}. So all we h a v e
to show is t h a t T'QT. Let R = (<<b}, <a, a}}}. T h e n /~ = R a n d rd(R,
M'~, M,) a n d rd(~, M',,, .M,,). ~ X ePot(M)r~C a n d Y' ePot(M',),
17' 9 ~ t h e n 17 = {{<a, a}}} a n d Y' = {{<b}}], t h u s :Y' 9
D10-3, 4) arc trivial, so we have T' ~T a n d hence T' ~oT because /~ ----R .
2.) <x, y} = @, r}
T a n d T' as in 1.) will do in this case, too. Obviously, -O(T'~T), Mp is
t , / !
a 2-0-matrix a n d M~, a 1-0-matrix. Suppose TrT' t h e n t h e r e is 0 : M , ~ 3 1 p ,
O(x, y) = <x} for <x, y} 9 M , . B u t t h e only <x, y} 9 Mp is <a, a} a n d
<a} r M~, i.e. -7(T~T').
3.) <x, y ) = <a, e}
Let M , = M , , = M: =
{f: R-+INIf(x) = const for all x 9 r: = Id,
C: = { ( f } l f 9 a n d I:----{f0} where fo(x)-= 1 for all x. Similarly,
= M~w = M': = M, r': = Id, C': = Pot(M,) a n d I ' : = M'. T h e n T
and T' are theory-elenwnts.
Generalized net structures of empirical theories. I I 181

LEMMA 1. .Not T@T'.


PROOF: S u p p o s e t h e r e is an R such t h a t RD(R, T, T'). Then b y D10-2)
(1) V ( X , X'} ePot(M2jp) xPot(M'~,p) (X' e A ( K ' ) A X e 9TtA
<X, X'} e ~--)-X e A (K)).
Take.f: R - + N , f(x) = 1 for all x. T h e n 0 =/=(f) ePot(Mpp). As rd(R, T, T')
t h e r e is a g e M~p such t h a t (f, g) e R. l % w let f ' : R-+N, f ' ( x ) -= 2 for
all x. Then X : = (f, f'} e Pot(Mw, ) a n d X e ~ . B u t X r C b y definition
of T a n d hence
(2) X 6 A ( K ) and X egJ~ and X ePot(Mpp).
Since rd(R, T, T') there is g' such t h a t <f', g'} 9 R. f =/:f' a n d rd(R, T, T')
yield g =/=g'. T a k e X ' : ---- (g, g'). Then b y definition of X '

(3) <x, x ' ) 9


Definition of T' yields V Y ( Y co_M'n--~y 9 Take X ' for 17 t h e n
(4) X' 9 and X' 9
B u t (2), (3) a n d (4) c o n t r a d i c t (1).
REMARK: FrOl]l T1 a n d Lenlnla 1 we o b t a i n : NotT~T'.
LE~,IMA 2. _Not T'@T

PRooF: Suppose T'QT. Then b y definition D9-3) Vx 9


9 I ((x, y} 9 F r o m this we o b t a i n t o g e t h e r w i t h D0-B5c): [I'l ~ III
w h e r e Ix] denotes t h e cardinality of x. B u t T' and T ~re so defined t h a t
III < [I'l, because I ' is infinite. This yields a contradiction.
:Now all we h a v e to show is t h a t TAT'. Clearly, Mpp = M~p, Mp = M'~,
M _ M ' , C _ C', I _ I ' a n d r = r ' .
REMARK: Frolll T1 a n d L e m m a 2 we o b t a i n : -Not T'oT.

4.) <x, y} ---- @, 0}

T a k e T a n d T' as in 3.) t h e n n o t TeT' a n d n o t T'eT. B u t TvT' with O:


0 (x): = x.

5.) (x, y) = (a, %~

L e t T b e a theory-element. Then TaT. Suppose T%T. M~ m u s t be an


k + m a n d a k + n - m a t r i x such that m < n which is impossible.

REMARK: This is t h e o n l y case where we reall), need v~, for, with


m = n in D6-1), we o b t a i n TaT'--->T~T'.
6.) <x, y) = <3, o}
182 Wolfqang .Balzer, Joseph D. S,ueed

Let, T a n d T ' be th(or;r n l e n t s such t,ha.t ' =I


~ !
n = {<xt, " * ' ~ ,% {a]-~
[<ak, . . . , x,,) e M v } ~nd M ' = { < . q , . . . , a , , , {a} i < x l , . . - , x , , ) eM}. Then
t
w i t h O: 3II~--+Ml, , O ( x i , . . . , x , , {a}) = <x~, . . . , x , " : T ' r T . B u t neit,her
M ' _ M n o r M _c M ' h e n c e n e i t h e r T ' a T n o r TAT'.

COROLLARY i . Th, ere are theor.q-ele,m,e,~t.ts T an,d T' su~;h that ,,eith.er
T~T' ,nor T'oT.
PROOF: T h i s follows imm~(lia.tely f r o m Lentm.a. [ a n d Lenmta. 2 of
t h e t,h e o r e m .

COROLLARY 2. Th, ere are lhe.r.q-elem, e.,ts T a,n,d T' such that ,.,either
T-OT' 9l.or T ' ~ T .

PROOF: This follows f r o m c o r o l l a r y 1 a n d T L

T4. I f N is a theory-,~t,et, N* the core-,,et a~xociated with N a~d N


the applieatio,rv-,n, et associated with, N the,., N * a,.,d N + are ~t,ot.s.

PI~ooF: VVe show t h e p r o p e r t i e s of D13-3) f o r A ~*.


(a) I f K e ]N*I t h e n h y definit,ion of A**, t h e r e is I such t h a t .~K, I ~ 'NI.
As N is a net,: ( K , I , .~ ,,K, II. ( D 1 3 - 3 a ) . S o K ~ * K . (b) li K<. K'
a n d .K' ~ * K " t,hen t h e r e aa'e I , I', I " s u c h that, <K, I } <~ (.K', I'~ a,n d
( K ' , I ' } ~< < K " , I " ) . F r o m D 1 3 - 3 b ) we o b t a i n .(K, I5 <..~I(_", I " ' , i.e.
K ~* K". (e) K , - ~ * K ' iff I( ~-I* K ' and K ' .i.~ K_ is t,rue 1)v. d e f i n i t i o n
of ~ * . N -~ is t,reat, ed simila.rly.

T5. I f N e ~J~ is a f i n i t e set of theor.q-eleme,,ts satisf!li;i, 9 D17-3) a tt,d

r = {<T', T" e (N :< N)IT'rT}


_--~ -- {<T', T) e(X• IT'rT a,t,d TrT'}
th,c.~, i N , r, ~ } is a theory-net
PIr 1.) Let, T E N a,nd O: 31~,-.-3I~, 1)e t,he i d e n t i t y o n J[~,.
T h e n T r T t)y D6).
2.) LL,t T " r T ' a n d T ' T T . W e show l h a t T " r T . B y D6), t h e m a t r i c e s
Of T " , T ' ~lld T are ],: @'~t,, 1,:-?'t~ a n d i ~ j, i ~ r resl)eet,i~'ely , w h e r e m ,:~ .tt,,
r~j and k _ m .
= i + j . I n a d d i t i o n t,here axe iunet,mns. . Or:. M p - , ,~-_]I'
~,,
t
O l ( . q , . . . , xk+,, ) = <,q, . . . , .r~+,,~'). = ,~,q, . . . , .ci.,.~' a n d O.z: 3I,,-+Mp,
O2(,i['l, . ,. , ; U. i + j ).: . <,921,
. . ,;Di!r)> Let. s : : ntill{]r ?-., t h e n T " a.nd T
ha.re s + u a n d s + v-ma.t,riees such that. v :-<., (s -i--.u, = 1,: + ~,, .~ + v --- i i-r).
t ,t
Let, O: M,-"~M~, b e d e f i n e d b y O = 0.,~:.0~. T h e n f o r x e M " : O~(.r) e M '
a.nd O.~iOt(x)) e M , h e n c e O(.c) e M . So we h ~ e p r o v e d Sha~ T " r T .

3) ,r =% y iff ( x r y a.nd vrx) b y d e f i n i t i o n oi -= ..

T6. I.['lV e 9J~ is a fi.t~,ite set of th.eory-ele.,te,,,ts satisfyh~g D17-3) a n d


a :: {.(T, T ' ) e ( N • N ) I T a T ' } the.a <~A', a, = } ,is a theo.ry-,t~et.
Generalized q~et structures of empirical theories. I f 183

PROOF: This follows immedia.~ely from D7).

T7. If N e 9)~ i.~ a .fbMte set of theory-elem, e~ts sati,~f!li'n,g D17-3) a,~d

o = {<T',T) e(N• -0. = { ( T ' , T > e ( N •

_0 {<T', T',, e ( N • N ) IT' a,.~d T arc o-equivale,~t}

=~- {.(T', T} e ( N x N ) I T ' a,~d T are ~-equivale.n,t}

the,~ ( N , 9, ~-ol; and .(.V, ~, -=~) are lheory-~,ets

PROOF: 1.) ] x ' t T e N a n d d e f i n e R : Mp->Mv, R: = Id. T h e n it is


e a s i l y c h e c k e d 1;hat R D ( R , T, T) a n d t h e r e f o r e T g T . I n case of 9 ta, ke
R : = I d on if m,"

2.) SUpl)OS(' T ~ T ' a n d T ' ~ T " . T h e n thel'e are R', R " such t h a t R D ( R ' ,
T, T ' ) a n d tt.D(R", T', T"). ],('t R be t h e l w o d u e t of /?' a n d R " , i.e.
I~ = {<x~ y)13z(.(,,'~ " e R ' A ~,z, y. e R")}. W e show t.hat R D ( R , T, T").
O b v i o u s l y , rd(R', T, I") and rd(R", T', T") i m p l y rd(R, T, T"). N o w let
] "It
,IX, X"> e Pot(Mm,) :.: P o t ( . Lm,) such t,hat, ( X , X " e ~ , X " e A (K") a.nd
X e %R. ( X , X " e~f~ imlflies tha.t, lqwre is a. f u n c t i o n e: X~-~X" wit, h
9kr, e(x)), e_~' f o r all .,'e .Y. T a k e X ' : := {zl >,, .... z'. e I~'A..I:, V(X)". 9 R " } a n d
let f ( x ) b e t h e m f i q u c l y d e t e r m i n e d z wit, h @', z e R ' A . I : , e ( x ) ) e R "
:fro" a' e X . T h e n <TX, .Y'/ e / ? ' wit.h f a n d .iX', X " . e / ~ " wit.h g: = col
N o w X " e . . l ( K " ) yivlds X ' e A ( K ' ) b y D9-2) a n d t,his X e A ( K ) for
~.he s a m e r e a s o n . So, since V.,' e I : l x ' e I ' ( . : x , x ' , ff_/::')AV;C' e I':]0"' ff I "
(>.c,' .~ " ' e R " ) implies V x e l q x " 9 I " t ,. ,' ,r. . , x ")." e R ) , we h a v e p r o v e d
I { D ( R , T , T " ) a n d l:h(,rvfore T g T " . S i n d l a r a rguntel~ts show t l m t t i m
l m , l w r t i e s of D10-1, 2, 4) are l;ransitivv in c:,.sr of ~o. So, fin' 0, it, r e m a i n s
to p r o v e (in l:lw a.1)ov,, n o t a t i o n ) : if ,/.e, ,,," e 1~, y " e M " :we suc.h t.hat
Y': v " ( y " ) ~,h('Jl tll(,r(' is // e 3[ sltc]l ~:lllt.~ Qt/, y " ' . e Ie a n d r ( / / ) - - , e .
JJel; Ix, x " ., e /) a n d .q " ~ M " w i t h ,,;. " = r " ( ! f ' ) l)e given. B y (lefinit&m
t t /_j R!lt >
of R, t.here is a n .,, such tha.t ,.a,' x"; e R ' a n d (:,,, , " e/~."
. . Now \.,,
^
9 R", !f' e Jl" and .,"' ;"(/1") y i e h t (D10-3) a n y' e 31' w i t h (9', :q" e R "
a lid r ' ( y ' ) ~ J,'. 8o we h a v e @;, .,;' : e R', ~.1' e M' a.nd r'(y') = x'. T h i s
a g a i n b y D10-3) gives :!n /1 e 31 such t h a t (y, y'). e R ' a n d r ( y ) .... x.
~-'N-O\V "if, ff'i @ ]/~' :llltrl (']1', i f " G J~'t llI(qtllS "~]l~t.~ r ff't.: ff 1~, ,SO WO ]I;IVO
f o u n d a n !1 e M such thai; Q/, .q" e / ? ttll([ f(.//) = .Y. This p r o v e s (TOT'A
A T'~oT"-.T gf").

3.) Front DI2-a, b) and t h e d e f i n i t i o n s of =o and --=7 we o b t a i n T =-e


iff (Ts, T' and T'c2T) a n d T ~ T ' iff (TOT' a n d I"OT) w h i c h eOn,ld(,t(,s
the proof.

4 - - Studia Logics, lj78


184 Wolfgang Balzer, Joseph D. Sneed

TS. I f N and N ' arc ~n~iq,~tely based a-theory-nets such that 93(N)
= ~3(N') = {<Ko, Io:'} a,~d ~ ' ~ . Y ' th,eu
(a,) ~(/V'*) ~_ ~[(~*)
(1)) A(3-'*) _ A ( 5 ~*)
I)ROOF:

LE~t~{A 1. I f zY = ([5rl, ~<, ~-;.. is a ~,et amt .re I.TVI then there is
some basic elem,e.~t z e I_Y! suck that x <~ z.
PROOF: :If ,~' is l)asic t h e n Lr.n,.ma 7k is r i ~ h t with z: -~ x. S u p p o s e
t h a t x is n o t ha.sic, i.e. ]y~(x ~< y~^ --l(x~y~)). :If y~ is basic ta.ke z: =: y~.
O t h e r w i s e *here is y2 such t h a t y~ <~ Y2^ 7 ( Y ~ Y 2 ) . As _~ is finite a n d
--I(y,:-~yj) implies t h a t Yi 9~ Y~ we c a n rel)ea* t h i s ])rocedm'(, o n l y a, f i n i t e
n u m b e r of t i m e s a.nd r a r r i v e at, .some y , w i t h

(1) Vu(y,~ ~ 'u-~?l, . . BX- constr~lction: ,~ -.~..?1~ ~% ... ~ y,,, h e n c e


. .~< u).
x ~< y,,. T a k e z: = y~ t h e n (1) sa,)'s tha.1; z is basic.

L E ~ . t A 2. I f Y is a theory-net with ~(~'Y) = {x} theft, y ~ x for all


y e INf.
Pnool~: I f y = x t h e n y ~ x t,3 definir of nets. O~herwise y is
n o t basic b e c a u s e x is t h e o n l y ])asic elentent of ~Y. F r o n t :iJenmm 1 we
obta.in a n z e ~ ( _ 5 T) w i t h y ~ < z . But, z e ~ ( N ) implies z ..... c, so y G x .
W e can n o w proYe t h e ~heorem. L e t .~ ~ 9A(N'*), t h a t is
(2) ~ ~ Posel(M~
(3) for all X e [~i: X ~ M
(4) t h e r e are a ' : IXI-+]X'*I a n d ~ _ (a' >~ ~') such t h a t
/ ~ t /
{4.1) _h*a: = - \ a , ~ , = : is a ad;heor3;-net , ~3(-Y,) = {:Ko, Io) }
(4.2) 5~ = 5"*
(4.3) tN+I = I.~I
(4.4) for all X , 1 ~ 1 . ~ [ : i; X~-/+ Y t h e n X _ ~ [I"
(4.5) for :ill Is e IX'*!: U { X i a ' ( X ) = K} ~ A ( K ) .
LEM_31-k :3. I f ~', -~Y', ~ a,~d a' are as i,~t, the theorem the~ there exists
a fuq~,ctio,~ S: a'-~(~' such that for all 1(, K ' e IN'*I and I, I ' e lY-l:

(1) i f S ( K , I) = ( K ' , I') the'~, K ' e IN*'~


(2) <K, I , <~x S ( K , I)
(3) -]~(K~,It\ ~'(K~IN*[^(K,I <t(K~,I~}:-~,
S ( K , I ) ^ ( K z , 1~', r ( K , I ) ^ < K ~ , I,., :/: S ( K , I))
(4) 'if ( K , I :, e a ' ^ \~K" , I' , ,. e a~ ' ^ ( K , I ; , / ~ , (K',I' <~
S ( K , I) then S ( K ' , I') = S ( K , I).

Pi~OOF: _~(St) = {(Ko, I o }, so b) * Lcnmm. 2 ~or a,ll ( K , I e 15"~1

(5) (K, I 4 , ( K 0 , I0} a,l~,d Ko e 15"*[


Generalized net structures of empirical theories. I I 185

because {<K0, Io>} = ~3 (N*), too. B y i n d u c t i o n on t h e n u m b e r of e l e m e n t s


in IN~t we p r o v e with (5) t h a t for a.ll <K, I> e IN~[ t h e r e is <K', I'> 9 Ihrl]
w i t h K ' 9 IN*l, <K, 1> ~<1 <K', I ' } a n d 7 3 < K 1 , Ix> 9 [Nil(K1 9 IN*[A
^ < K , I> ~<~ <K~, I~> ~<~ <K', I ' } ^ <K, I> # <K~, I~> # <K', I'}). W e pick
o u t exactly one such <K', I'> for each <K, I> a n d call it S' (K, I). T h e n S':
a'-+h' satisfies ( 1 ) - (3) above. I n order to get condition (4) we h a v e t o
change S' in t h e following way. L e t us say t h a t x is t h e i m m e d i a t e pre-
decessor of y ff x, y 9 ]N~[ a n d x ~<~ y A 7 3 Z 9 [N~[ (x <~1 z <~1 y A X :/: z :/: y).
T a k e N3 to be t h e set of a.ll x 9 IN~I t h a t a.re i m m e d i a t e predecessors of
s (x). T h e n S' restricted to N3 satisfies (4) above. Now let x 9 Na a n d y
be an i m m e d i a t e predecessor of x. ~ 7 ( x <~1 S'(y)) we set S ( y ) : = S'(y),
if x ~ S'(y) we set S'(y): = S(x). Obviously, all elements of IN~[ are
in t h e d o m a i n of S a.nd S satisfies conditions (1)--(4) of t h e L e m m a .
N o w we define a: IXI-+[N*I b y
K(I) if a ' ( I ) 9 IN*I
a(I) : = if a'(I) r IN*l a n d t h e r e is 1'
with <K, I ' } = S ( a ' ( I ) , I).
F u r t h e r , ~2 ~ (a • a) will be defined b y <K, 1> ~<2 <K', I ' } if a n d only
if one 02 t h e following conditions is sa.tisfied:
(i) a'(I) 9 IN*lA a'(I') 9 IN*lA <K, I> ~<~ <K', I ' }
(it) a'(I) 9 I-N*i^a'(I') r [N*IA(K,I><<.~(a'(I'),I'>
(iii) a'(1) r I-N*IA a ' ( I ' ) 9 IN*IA S ( a ' ( / ) , I) ~<x <K', Z'>
(i'v) a'(I) r I-N*IA a ' ( / ' ) r IN*IA
[ 7 ] I ~ ( a ' ( I ~ ) 9 IN*IA <a'(I), I> ~<x <a'(/t), Ix>
4, r) =
A ( a ' ( / ) , 1), <~ (a'(I'), I ' > v
:tI~(a'(I~) 9 [N*[A (K, 1> <~ <a'(I~), I~> ~.,
<K', r>)].
There is no circularity in this definition. The use of ~<~ in clause (iv) is
short h a n d n o t a t i o n for repeating ela.uses (it) a n d (iii). We ca.n n o w show
tha,t
(a) N2: = <a, ~<.,, =), is a a-theory-net, ~(N2) = {<Ko, Io>}
(b) N~ -- _N*
(c)
(d) for X , Y 9 IXI: X ~<+ :Y->X ___ :Y
(e) for K e IN*l: U { X l a ( X ) = K} 9
From conditions ( a ) - (e) t h e t h e o r e m cfl,n be p r o v e d w i t h o u t difficlfl~y.
(a.) AL = <a, ~<2, = > is a a-theory-net with $(N~) = {<Ko, I0>}.
Obviously, a is a finite set of t h e o r y - e l e m e n t s a n d ~2 -- (5 • a). D17-3)
is satisfied since [N2I = 5.
(a.1) x 4 fi for x 9 ~.
186 Wolfgang Balzer, Joseph D. Sneed

L e t x = ( K , I ) e a be given. If K 9 IN*I t h e n x e ~' a.nd x ~<t x beca.nse N~


is a n e t . So x ~<~ x by- definition of ~<2. If K r [N*] t h e n (a'(I), I> <~, ,,a (I),
I}AS(a'(I), I) = S(a'(I), I)A -]~I,(a'(I~) e IN* I A (a'(I), I> <~, (a'(IO, I,>
<~ (a'(1),I>) for or I<~i ~ I~<~ + I, so I _c I , _c I b y (4.4), i.e.
I = I~ a n d therefore a'(I) = a'(IO which is impossible in view of ( ; ( I )
r IN*l. So 1)y ela.use (iv) of t h e definition of ~<2, a: ~., .~.

(a2) x ~<2 Y ~<~ z implies x ~<~ z for x = <K, I>, y = ( K ' , I ' a.nd
z = <K", I"))e a. W e h a v e to consider eigh~ c~ses in v i e w of
t h e definition of ~2. L~t us consider two of ~hem. First, let a'(I) 9 IN*I,
a'(I') a n d a'(I") r ]N*I. So <K, I> ~ (a'(I'), I'} a n d either -]][~ (a'(I~)
9 I-Y*lA ( a ' ( F ) , F> ~<~ ( a ' ( I , ) , I,> ~., ( a ' ( I " ) , I")>)AS (a'(I'), I') = S ( a ' ( I " ) ,
I")A<a'(I'), I ' ) ~ , <a'(I"), I " ) or 3[~(a'(I1) 9 ]N*IA(K', I'> ~<., ~a'(I O,
I~) ~<o ( K " , I " ) ) b y de[inition of ~<~. I n the fh'st c-~s:~' we o b t M n
<K, I), <~l (a'(I'), I') <~ <a'(I"), I") s:nc~ N~ is ,~ n:r a,nd so (K, I , ~2
( K " , I " ) b y (ii) of the definition o~ ~<~. In t'.,lo secon4 e:~v>, ,[K', I",
~<~ <a'(I,), Iz) implies (by (iii)) SIla'(I'), I') ~.t (a'(/0, /*1> ~][1([ ( a ' ( I t ) , It',
~2 ( K " , I " > implies (by (ii)) (a'([O, I~), ~ t ( a ' ( U ' b I">. F r o m this we
obt, Mn with (2) of Lemma. 3 a.nd the, t r a n s i t i v i t y in ,'Vt: <K, ['~ ~ t .(a'(I")
I " ) t h a t is .v ~<.. z. The second c~se we are going ~o consider is t!u~ ~,)llowing.
a'(I), a'(I') mid a'(I") r IN*I a n d b o t h ~<2-relations hold bat:rose oi~ t h e
fh'st clans(, of (ix,), i.e. --13It(a'([t) e IN*lA(a'(I), I> <~ ( a ' ( / O , l~)
r:>)^6'(c(r), ' .,a(I),' I'',A
A--]:tI~(a'(I.,.)elN*fA,at,
f ' ~ " ~ I, I'~ ~t<a ' (Io), I.,'><~<a'(I"),
"*" / I" )A,S(a
~ '

(I"), I") = S(a'(I'), I')A(a'(I'), I') <~ (a'(I"), I " ) Suppose t,h:,~:

(6) 313(a'(Ia) e ]N*[A @'(I), I} ~<, ,(a'(Ia), Ia> ~ (a'(I"), I " ) .

Take (K4, I~). : = <a'(Ia), Ia) t h e n

(7) ./,K4, I[~ # <a'(I), I \ bec~mse K4 e IN*I and a ' ( [ ) r [N*I.

(8) .(K~, I4) # S ( a ' ( I ) , I) for otherwise (K~, I4, ~ t (a'(I"), I'",
~ (K4, I4)' b y (6) a n d t r a n s i t i v i t y in B r . B u t this w o u l d imply t h a t
I " = I4, hence I(~ = a'(I") which eonfradiets K 4 e IN*I. F u r t h e r

(9) <a'(I), I} <~: (K4, 14'.> 4 : S(a'(I), I) b y (6), 2) of Lemma, 3 a,n(l


H(a'(I"), I " ) = S ( a ' ( I ) , I). At the other ha.nd 3) of Lenmu, 3 oives us

(lo) -13 LI z h'(K e IN*t^ I> <IG,I;, s(.'(z), i)^


^ <.'(I), I> # # 6'(.'(I), I))

which is ~ contradiction to (7)--(9). So a.ssuntption (6) is fa.lse and for


x ~ 2 z it stfffices t~o show t h a t H(a'(I), I) = S(a'(I"), I " ) A ( a ' ( I ) , I "~-1
< a ' ( I " ) , I " ) b u t this follows fl'om our assmnl)tions on t h e case considered.
Ge~eralized net structures of e~npirical theories. II 187

The r e m a i n i n g cases of a2) are t r e a t e d similarly.


(a3) x = y~x <~2 y A y <~ ~x for x, y e INd.
The direction f r o m left to right is trivi,~l in view of (al). For t h e converse
direction we h a v e to consider four cases according to t h e definition of ~<~.
F r o m these, t h e only c o m p l i c a t e d cases is t h e following, a' (I), a'(I') ~ IN*I,
where x = <K, I ) a n d y = <K', I ' ) ,%nd b o t h ~<2-relatious hold because
of t h e second clause of (iv) in t h e definition of ~<~, i.e. there are I x and I~.
with a'(I1), a'(I2) e IN*] a n d x <~ <a'(Ix) , Ix} <~ y ~nd y <~. <a'([.,), ~2>
< ~ x . F r o m this we get S(a'(I), I) <~ <^'(Ix), Iz> <~ <a'(I'), I'} and
S(a'(I'), I ' ) ~ <a'(I~), A.> ~ <a'(I), I> b y definition of ~2. W i t h t h e
t r a n s i t i v i t y of N~ a n d (2) of L e m m a 3 we obtain I ~ + I ' a n d I ' ~ + I a n d
so f r o m (4.4): I = I'. B u t t h e n K = a'(I) = a'(I') = K', so x = y. The
r e m a i n i n g cases are t r e a t e d similarly. So N2 is a-theory-net. N e x t we
show
(ad) ! ~ ( N 2 ) = {<Ko, I o } }

a'(Io) = K0 e [N*] because of


(II) {<K0, I0>} = ~3(Nx) = ~3(N).
So a(Io) = Ko b y definition of a, t h a t is, <K~, I0> e IN.~I. Let <K, I>
e IN.~I. If a'(I) 9 IN*I t h e n by (11) a.nd L e m m a 2 <K, I> <x <Ko, Io) a n d
therefore <K, I> <~ <Ko, Io>. If a ' ( I ) e IN*I a n d S(a'(I), I) = <K~, I~>
15~xl t h e n l~y (11) a n d Lemma. 2 <Kx, I D ~<, <Ko, Io>- So <K, I> ~<~ <Ko,
Io'. b y (iii) of t h e d(.finition of ~<~. F r o m this we obtain Vx e lNd (<Ko, Io>
~23'--><Ko, Io> ~ x <2 < K o Io>) a n d so x = <Ko, Io> b y (a3). That is,
<Ko, Io."" z ~ ( N D . Now let y e ~(N~). T h e n f r o m <Ko, Io> e ~ ( N ~ ) we get
Y ~<2 <Ko, Io> and f r o m y e ~ ( N 2 ) : <Ko, Io> ~<2 Y. So b y (a3): <Ko, Io> = Y-

(bl) IN*] = ]N*[. This follows fa'om (4.2) and t h e definition of a.


(1)2) for K , K' e IN*[: K ~* K ' ~ K ~* 1(_'
The dh'ection f r o m r i g h t to left follows from N [ - N ' , (4.2) ~nd the de-
finition of ~2. F o r t h e converse direction t h e following lemm~ is useful.
LEM~I 4. I f K , K ' e ]N*I a~7,d K <~*K' the~ K <~*K'
PROOf,': (4.2) and N ~ N'.
~ o w for t h e r e m a i n i n g p a r t of (b2) we consider K, K' e IN'] with
K ~<* K'. T h e n there are [, / ' e ]51 such t h a t <K, I} <~2 <K', I'>. According
to t h e definition of ~<~ we h a v e to consider fern' cases. Let for instance
a'(I), a'(f') ~ I-N*I. First if '~]I~(a'(fx) e ]~*]^<a'(I), I> <~ <a'(I~),
1,>~<~ <a'(/'), l'>)^S(a'(1),I)= N(c,'(I'),l')^<a'(1),1} <z <a'(I'),1')
t h e n b y definition of a t h e r e are Ie, 13 such t h a t ~%'(a'(I), I) ---- <K, Ie> and
188 Wolfgang Balzer, Joseph D. S~reed

S (a'(I') I') = <K', I3>. So <K, I~> ---- <K', I3>, i.e. K -- K'. As <K, I2> e ~',
we infer <K, I2> ~<1 <K, I~> f r o m N, being ~ net. B u t this means t h a t
K~<~ K ' hence K~<* K ' b y Lemm.~ 4. The second possible case is
tha.t 3Ii(a'(I1) e IN*In <K, I} ~<. <a'(I~), I1> 42 <K', I'>). I t follows t h a t
K <~* a'(Ix) a.nd a'(I1) <~* K'. I n b o t h c~ses it is e~sily shown th.~t b2)
holds, so K<~* a'(I~) ~nd a'(I1)<~* K'. :Now K~<* K ' follows f r o m Td.
The o t h e r ca.sos a.re t r e a t e d simila.rly.
(b3) for K, K' e IN2]: K ~ K ' iff K ~ * K ' .
This is trivial because ~ a.s well as ~ * is t h e identi~y.

(c) I~+l = I~l.


The inclusion f r o m left to right follows th'om t h e fact t h n t Dz(a) -- t-~!.
Conversely, if I e I:~1 t h e n I e IN+I becnuse IXI is t.he d o m a i n of a'. So
~,here is some K such t h a t <K, I> e a'. If a' (I) z IN*I t h e n <K, 1> z h and
t h e r e f o r e I e I~+1. If a'(I) ~ ]N*I t h e n t h e r e are K', I ' with S(a'(I), I)
= <K', I'>. B u t this menns t h a t <K', I> z ~, i.e. I z IN+I
(d) for X, Y e I~1: if X <~+ Y then X ~_ Y.
If X ~< + X~ t h e n th(.re nre K, K ' such tha.t <K, X> ~<~ <K', 17>. Suppose
a'(X) e l-N*] a.nd a'(Y) ~ IN*I. T h e n <K, X> <~, <a'(Y), Y> b y definition
of ~.2. This implies X <~ + Y, hence X __q Y b y (4.4). The remaining" eft.sos
~re trOd.ted similarly.
(e) for all K e I N * [ : U{Xla(X) =K}eA(K)
:Let KeiZT*I a n d X e l E ] such t h a t a(X) = K .
If a'(X)r then
~(x) = .'(x), so x e { X l ~ ' ( x ) = K} therefore X ~_ U { X / a ' ( X )
and
= K} = : P. If a'(X) ~ 1I*1 t h e n t h e r e is I ' such t h a t <K, I'> = S(a'(X),
X). So <a'(X), X> ~<~ <K, I'> b y 2) of :Lemma 3 a n d f r o m this X ~< + I ' .
(4.4) 5~ields
(12) X _~ I'
As <K, I'> e ~t': K = a'(I') a.nd therefore I ' _q P. This t o g e t h e r with (12)
gives X _~ P . As X was a.rbitra.~ T we h a v e p r o v e d X _~ P for all X with
a(X) = K. So U {x[ a(X) = K} ~ P. ]~ut _P e A ( K ) b y (~.5). This yields
U { X I a ( X ) = K} e A (K) b y tra.nsitivity of constraints. TS-b) is a trivial
consequence of T8-~) a.nd D23).
TP.I f T' and T are theory-elements and RD(R, T', T) (RD(R, T', T))
then for all TI~rT' there is some T2gT such that l~D(R, T1, T.~) (RD(R,
T~ , T~))
P R o o f : :Let us first consider RD. Suppose we h a v e T, T', T1 as in
t h e theorem. T h e n we define T~ a.s follows. M2: = My, M~v: = Mvv,
~'~: = r, M~: -= S ( M ~ ) n M ~ M + where S(I~): = {~ e MvI~y e :Y(<y, z>
Generalized net structures of empirical theories. I I 189

~ 3 , ~ e M1 (x~ -- r~(,,~)^ ( , , .~" e n))}, X e C ~ X e CA 3 I ~ e C , ( X _= S ( Y ) ) ~


v3,~ e M ~ ( X = {x}) .qnd [o-: = Z.
] 2
LE,~[~IA 1. :To-: = \ 9,/:, lur 2 ~ p , :lj,, r2, M._~ C2~,
~\
I2) is a theory-element.
]?ROOF: :]I~, is a. n m.trix, M~p = g~(M~,) a n d M2 ~- 3I~,. C2 is a, const, raint,
for M~:2
(a) O r C~ boca.use O r C.
(b) C2 ~ O because 2 i v #- O.
(c) L e t X e C : a.nd We~.~, W ~ _ X . Then W e C A W ~ _ X ~ _ S ( Y ) or
W = X. I n n n y c~se W e C~, so C2 is tra, nsit,ive.
(d) for all x e M~: { x ) e C2 b y definition of C~.
o / 2
Finally, I . , _ M]~p b y dcfinit,ion of I., a,nd M~, beeausc T is a t,hcory-
-elenmnt. This proves Lenum~ 1.
Ll~.~nt 2. R D ( R , T~, To)
P~OOF: (a) rd(R, (I~,,
]' ~ M~,)
~ because M~ = :l/~, ' M;, = M ; and
rd(R, ME,' Mp), a.nd rd(R, ;lily , ;ll~,~)
" beca.use ~]g~
' ~ i ; , , , , '~
a n d rd(R, "
(b) V(x~, x~} e M'.. "<M~,p((x~, x2) e k-->Vz e M.~-~zx e MI (x~ = rz(z)

PI~OOI~: If <x~, xo-} e / ~ a.nd z e Me- is such ~h~t x-a = re-(z) ~hen b y
definition of M + t,here exist,s an zt e M1 such ~hat, x, = r , ( z , ) n (z,, z} e R.
(c) V ( X , :Y} e P o t ( M ' ~ ) • e P o t ( M 2 ) n C ~ A X e 0Jr A ( X ,
:Y~, e f~-+X e Pot(M~)nC~)
P ~ o o ~ : (i) X _ M~
Suppose I x e X ( x r M~). F r o m ( X , Y} e / ~ we haa, e 3Y e I7 ((x, y} e R).
F r o m :Y ~_ M~ we ge~ y e M~. B y definition of M2 a n d S(M~) ~here is
z e Mx such tha.t, (z, y} e R. Now @, y} e RA (z, y ) e R implies x = z e M1
b y DO-B5e). This is a contradiction.

(ii) X e C1
]~ e co y i d d s 3 Z e C',l Y ~_ S(Z)) ol. 3x e M ~ ( Y = (x}). F~'st,, let, Y = {x}.
T h e n X = {y} ( D 0 - B 6 ) w i t h y e M~, a n d h e n c e X e C 1 S e c o n d , let, :Y
be s u c h that, ~Z e CI(Y ~ S(Z)). W e s h o w t,ha.t, X ___ Z. Let, us ,~ssmnc
t,hnt =Ix e X ( x ~ Z). F r o m x e X a n d ( X , Y ) e k w e get, 3y e Y((x, y) e R).
Y ~_ S(Z) y i e l d s 3z e Z ( ( z , y) eR). But, (x, y) e R ^ ( z , y) e.R, implies
x = z e Z in cont,r,~dict,ion t,o mtr ,~ssump~ion. So X g Z. But, Z e C~ a.nd
t,herefore X e C1 ( D 0 - B 3 c ) .

(d) Vx e I~3y ~ I2(@, Y) e -s


190 ll'olfga,ng Balzer, Joseph D. S,ueed

This follows fi'om I~ ~ I ' a n 4 t h e d(-fmition of i e . So we h a v e p r o v e d


r t h e o r e m f o r RD.
L e t us n o w ~ u ' n ~o tlie ease t~D w h i c h t m ' n s o u t t o b e m o r e c o m p l i -
ca$,ed. L e t T, T', T 1 l)e ~.heory-elenwnts such t h a t R D ( R , T', T) and T~ aT'.
1.) There are .fu.,wtio.,,.~ .f: 3ljw-~M~, , f': :l/j,~)-,:ll~) such that Vx, y ~ Mm,
(a, :- ;,l~f(x) ~ .f(y)), V.r, 71 ~ M,,,, f,r =A ?/~Tf (.r.) ~ f (y)) aq,.d r(f(x))
-_ .r, r ' ( f ' ( x ) ) = x. T h i s is easily s h o w n b y usin,'., t h e a x i o m o:[ c h o i c e .
So we h a v e

w = 4
Let S(I'): = { x e Mm, l~.q 9 I ' ( Q / , x e R ) } . N o w T.> will be d e f i n e d 1)y
= :v,,,>, x 9 c+-+x
A3Y eC,(X c_.f(S(f'(Y>)))v3x9 = {x})a.nd I~: = I.
3.) I t is easily c h e c k e d iliad5 T e iS fl t, h e o r y - e l ( q l t e l l t a n d TeaT.
4.) N o w we s h o w t h a t R D ( R , T1, T.,).
(a) rd(R, Mm>
' ~ , Mju,)
" follows :~l'Olll. t]l(' d e f i n i t i o n of a.
(b) V ( X , Y . 9 , >;I'ot(Mm,) I Y 9 A(K.>)AX 9 gJIA ~~X, i~, _~
=.X 9 A (I(_~))
r
PgOOF: Le~: ] / ~ A ( K ~ ) , X c Mm, 1)e g i v e n such t.hat ( X , ]$) ~ k .
We h a v e t o s h o w that, X e A(K~), i.e.
(1) :Iz(~'(z) = x^z es)ot(]C)ncd.
;Let us d e f i n e Z : = if(X).
W e claim *hal:
(2) ~'(Z) = X

I>ir If .,, 9 c .Ira, t h e n . f ' ( . r ) 9 JI~, a n d r ' ( f ' ( x ) ) x bY 1.). So


=tz: = i f ( x ) 9 x), i.e. X ~ Y(Z). Con~wrsely, if . v e t ' ( Z ) t.hen
3.u +Z(r'(u) = :r)..u +Z ilnplies 3.V =if(v)).9 So we ha.~-e r'(u)
.... r ^ r ' ( + 0 = r'(.f'(v)) = v, w h e n c e x = v e X .
(3) R(YnS(X)) = X ( N o t e tiu~t /) is a. fm~ction b y D O - B S c ) .

PUOOF: Lot.ceX, thenl)y ( X , Y ) e / ~ we h a v e : l y 9 eR).


So qy 9 Y n S ( X ) : I~.(.,/) --: x, i.e. x 9 Con-versely, if z 9 _#~( Y n
ns(x)) the~, 3.,,. 9 ~ ' n S ( : r ) ( ~ = ~(.,,.)), ~.e. ,~~ , u,) 9 R . u 9 S(X) imlflics
=IV E X ( ( ' 0 , '~/)'. 9 R). F r o i l t z , it". 9 R A ~r, n . 9 ]{ we get, : = v 9 X .
(4) Z ~ =11l

PROOF: L e t 3~ ~ Z.
J.) x ~ Z A , r ~ X ' ( X ) -- f'(i~(YnS(X)))
2) x 9 Z-->~7I 9 R ( I ' n S ( X ) ) ( x -- f ' < y ) )
~) y 9 ~ r n , s ' ( x ) ( ~ m , ~;, 9
4) z 9 ]'->::tb 9 Jl2(r.2(b) -- Z) b o c a r i s e ~." 9 : ( ( K 2 )
Generalized 'net structures of empirical theories. I I 191

5) z e 8(X)-~3u e X(<u, z> e R)


6) b e M2-+3v e S ( f ' ( M , ) ) ( b = f(v)) Def. of M.,
7) v ~ ~ ( f ' ( M 1 ) ) - + 3 w e f ' ( M i ) ( < w , v) e R)
s) w e f ' ( i l ) - + 3 a e M l ( w = f'(a))
9) v = r2(f(v)) = r2(b) = z 4)~ 6)
10) i f ( y ) = x-->y = f ' ( f ' ( y ) ) = ]'(x) 1.)
11) v = j'(x) 2), lo)
12) <y,z>eRA<W,v>eRAz =v 3), 7), 9)
13) w = y 12)
14) f'(a) = f ' ( x ) 13), 11), 8)
15) a = x 14) because of 1.)
i 1 b y 15) a n d 8) which proves (4).
(5) ZeC~

PimoF: 1) Y e A ( K 2 ) ~ 3 U(Y(U) == Y ^ U ePot(M~)nC.~)


2)
3) U = (x}-+3y e M;,(Z = {y}) D0-B6)
4) U = {x}-+Z ~ C~ D0-B3d)
5) y eZ->3z eR(YnS(X))(f'(z) = y) this is 2) of (4)
6) z ~ / ~ ( y c ~ s ( x ) ) - + 3 u ~ I~c~s(x)(<~, ~<> ~R)
7) u e S(X)~3v e X ( < v , ~G e R)
8) r'(y) = r ' ( i f ( z ) ) = z e X 5), 1.), (2)
9) z 9 X - > 3 a e :Y(<z, a~ e R) bec,~use <X, Y> e / ~
10) i:(U) = Y A a e Y ~ 3 b e U(r(b) = a)
11) -]=lx(U = {x}-+b e U--+:le e S { i ' ( V ) ) ( b =: f(e)) 2)
12) e ~ S(f'(V))-->3d ~ f~(V) (<el, e) ~ R )
13)
14) a = r(b) = ~'(f(c)) = c 10), 11), aria 1.)
15) if(z) = yA<z, a> e R 5), 9)
16) z - - ] ' ( i f ( z ) ) = f ' ( y ) 15) n,ml 1.)
17) (d, e) e R ^ (z, a) e / ~ ^ a = e 12), 14), 15)
18) d = z 17)
19) f'(e) = f ' ( y ) 16), 18), 13)
20) e = y 19)
21) y e V 20), 13)
22) 7 3 x ( U = { x } ) - + z _ v 5)-21)
23) --13x(U = {a:})-+V 9 C~ 2)
24) 7 3 x ( g = {.~,})~z e c, 22), 2a), D0-B3c)
25) Z e C~ 24), 4) This proves (5).
192 Wolfgang Balzer, Joseph D. &need

(6) for all .,, 9 I1 there is :,! 9 I2 such that <x, y} e R.


If x e I1 t h e n x 9 I'. Now RD(R, T', T) implies 3y 9 I = Io-(<x, y', 9 R).
This completes t h e proof of TP.
T10.There exist theory-eleme,tt,ts T' and T and R such that RD(R, T', T)
(RD(R, T', T)) a,~d for some TlathT' there is ~o T2alnT such that RD(R,
To) T1,
PROOF: :Let a, b 9 9J~. i r e write " a " for "<a>" a.nd % " for "(b ;' ? 9

:Let T' = Tt: = <<{a, b}, {a b}, Id, {a, b}, {{a}, {b}, {a, b}}}, {a, b}> a n d
T: = <<{a, b}, {a, b}, Id, {a, b}, {{a}, {b}]), {a, bi>. I t is easily sho,wn Om.t
with /~: = Id: RD(R, T', T). B y T6): TlcrT'. :Let G: = Mp in D8) t h e n
t
T~ art,T' because r ' (C') = Pot (Mvv) \ {O}. Now suppose ~ Tz ~u, T such tlta~
RD(R, T~, To). T h e n
(1) 2 = IM~,I < [M~I a n d (70 ~ C
Three ca.ses are possible.
1) Co = C. T h e n r.,(C=) r Pot(M;w)\{O }.
2) Co = {{a}}. T h e n M~ = {a} because r~(Ce) = P o t ( M ~ , ) \ { O } .
B u t t h e n IM~I < IM,~,I in contradiction ~o (1).
3) C = - {{b}] leads to a c o n t r a d i c t i o n in t h e same w a y as 2).
8o in each case ore" assumption is false.

Tll. (a) /~or all R 9 ~ , if T a~d T' are K-theories a.~d P - ~ R D


(1~, T', T) then R stro.~gly completely reduces T' to T.
(b) T l l - a ) becomes false whe.~ a is replaced by au,

PrcooF: (a) If P--~R~(R,T',T) then {R~iR, hr~,hrp) D(26-a), so


D26-bl) is ~rue. I t remains to slmw t h a t for all h r' 9 Jf~' t h e r e is some
N e W such t h a t ~R~(R, N', N). :Let N' 9 i.e. N ' is a a-theory-net,
t / t t !
based on To: = \Ko, I'0> where <Ko, I~,> is t h e b~sic e l e m e n t of N~,
t t I
I , c I0 a n d Np I N ' . L.~,t To: = <K0, Io> be t h e basic e l e m e n t of Nj~.
~ o w t h e desh'ea N is c o n s t r u c t e d %s follows.

1) I f I~: = I o w { x e M j w l 3 y 9 1 4 9 a~d T~: -=- <Ko, I~> the~


2~2)(R, To, ~).
PROOF: lq~f0111 iR~(R, N', 2r we obtain RD(R, <K'0, I'p>, To). The
proof of T9) works essentially with cores a n d remains valid w h e n t h e
i n t e n d e 4 applications of b o t h specializations are c h a n g e d p r o v i d e d only
t h a t D10-4) remains Vrue. So we can substitute I'o for I', and T 1 for To
a n d still ttD (R, To, T1) is ~'ue. F o r t h e cores a m o n g which R works r e m a i n
t h e same a.nd b y definition of 11 D10-4) is satisfied. We t a k e T1 to be
t h e basic element of _hr. Obviously, Io c_c_I~.
Generalized net structures oJ" empirical theories. II 193

! r

2.) I f T' e IN'Ic~ INpl the~, by ~79 (R, N~,, Np) and D25-2) there ,is T2 e IN~,I
with R D ( R , T', T2). All these T2's are members of N.
r t i
3.) I f T ' e I N ' I \ ] N ~ , I the~ T'aTo since {T0} = ~ ( N ' ) and N~,ZN'.
F r o m T9) we obta.in ~ To-aT 1 with R D ( R , T', To). All these T2's ,~re m e m -
bers of N, too.
4.) Finally, a.ll m e m b e r s of N v ~re in N. All T e INI a.re specializ~tions
of T 1 for those o b t n i n e d b y 2.) or 4.) axe 9 INpl a.nd so TaToaT 1 a n d those
obt~ined b y 3.) are specia.liza~ions of T~ b y construction. Therefore INI
p r o v i d e d with t h e a-relation yields ~ a-theory-ne~ N = <INI, a, = } bused
on T 1. B y 4.): N~ _~ N, a d d ~ ( R , N', N) is s~tisfied becsuse of t h e
c o n s t r u c t i o n of N in 1.) - - 3 . ) .
! !
(b) I f T ' : = G,:~, I~,, N;;, .*"> is such tha~ <Ko, I v > = T' where T'
is a.s in t h e proof of T10), [Nj, I consis% of T' only; if fm'ther T: = <K0,
Ip, ~ , W> is such t h a t <Ko, Iv} = T fl'om T10) a.nd Nv consists of T
onl 7 t h e n p _ ~ R ~ ( R , T', T) where R is as in T10). For R D ( R , T', T)
implies ~R~ (R, .N],, _A~)b y definition of _Y~'~a n d 5~,. B u t R does no~ sta'ongly
c o m p l e t e l y reduce T ' to T. F o r if Tz is t h e specia.liza.tion of T' in t h e proof
of T10) a n d 2V' is defined to contain just T' a n d T~ t h e n ~u e W'. Bu~
b y T10) there is no T2cr~hT with R D ( R , Tz, To-) hence there is no N e~V
wi*h 9 ~ (R, .N', N).
T12. (a.) I f S e 9~ is a finite set of K-theories and
< = {<x, y> ~ S • S l 3 n ( ~ - - ~ V ( R , x, y))}

~ ' = {<x, y> e S • S Ix ~<' y a~uZ y ~<' x}


theq~ <S, <~, ~ } and <S, <~', ~ ' ) are q~ets.
(b) Let S ---- (T~, ..., T,,} be a set of K-theories a~d {T i} = ~(N~) for
i ~ n. I f <{Tr ~< n,}, ~, ----> is a a-theory-net and
< = {<T;, T~> e S • SIT,~T~}
:: {<Ti, Ti} e S • SIT, aT i and TI~T,}
~..

then <S, <~, ,~) is a net.


P~0OF: (s) Obviously P - - ~R?D(Id, x, x) ~nd P - - ~R~D(Id, x, x) so
x <~ x und x ~<' x for x e S. We h~ve ~o show t h s t P - - 9~:D(R~, x, y) and
P - - 9~D(R~, y, z) imply t)--9~D(R, x, z) where R is suit,~bly chosen. The
~'a.nsitivity of (strong) r e d u c t i o n of t h e busic elements of f o r m <K0, I~>
is gu~ra.nteed b y T7). This implies ~ra.nsitivity of the p~a.~digm nets s n d
even of t h e "possible" expa.nsion sets ~V.
(b) T~aTi (T6) implies T I ~ < T , ~nd from T ~ < T i~<T~ we
obt,%in T i a T i a T ~ hence T~aT~ (T6) nnd T~ ~< T~. Finally, Ti,-~T i iff T~aTi
~nd T~ aT i iff T i <~ Tj und T~ ~< T~.
194 Wolfgang Balzer, Joseph D. SneeeI

Errata to Part I
Definition D16 (Studi~ Logic~ Vol. 36 :No 3 (1977), p. 29) should be
as follows :
D16. If ~ a n d N' ,~re nets a.nd N ' K _~r, t h e n _N' is a,n initial p a ~
of _~ iff for all x e (15~]\ I-~'l) ,qnd all y e IIY'l: not, y ~ x .

Bibliography
[1] T. S. Kuilx, T h e S t r u c t u r e o[ ScientiIic R e v o l u t i o n s , Second edition, Chicagr
1970.
[2] C. U. I~IOULI~I~S,A logical reconstruction of simple equilibrium thermody.~amics,
E r k e n n t n i s 9 (1975).
[3] J. D. SN~.~D, T h e logical S t r u c t u r e o~ M a t h e m a t i c a l P h y s i c s , Dordrecht 1971.
[4] J. D. SNOOP, Philosophical problems in the empirical science of science ." a forma~
approach, E r k e n n t n i s 10/2 (1976).
[5] W. ST~G.~t0r.LER, T h e o r i e u n d E r I a h r u n g , Zweiter Halbband, Berlin-Heidelberg-
-New York 1973.
[6] W. STV.G~IOLLER, Accidental ("non-substantial") theory change and theory dislod-
gement : to what logic can contribute to a better undcrstanding of certain phenomener
in the dynamics of theories, E r k e n n t n i s 10/2 (1976).

S E M I N A R FOR I ) t I 1 L O S O P t t l E TEcn~1scmt HOGESCIIOOL


I , OGIK UND W I S S E N S C I I A F T S T H E O R I E :E1NDHOYEN/HOLLAND
U N I V E R S I T X T ~/[ fd N CII E N

Allatum est die 27 Octobris 1976

8ludia Logica X X X V I I , 2

You might also like