ISTV£N B_.~,TORI 
WORKING WITH THE INTERACTIVE VERSION OF THE 
T.G.T.-SYSTEM OF JOYCE FRIEDMAN 
The present paper does not claim to be a description of the TGT- 
System, since it was already presented by Professor Friedman herself 
at the International Conference on Computational Linguistics in Stock- 
holm in 1969. In addition the system has been described also in the 
book Jo'i'cE FRIEDMAN, A Computational Model of Transformational Gram- 
mar, Elsevier, 1971. Our intention is to present the new interactive ver- 
sion of the TGT-System, which has been developed at the Basic Re- 
search of I13M Germany, and to show how it can be used in linguistic 
research. 
In order to appreciate the present interactive version, it will be, 
however, necessary to recall some essential aspects of the TGT-System, 
yet we do not want to discuss the Friedman System as such in a 
systematic fashion. 
Accordingly, in the first part of the paper I shall talk about the 
batch version, and about our experiences with the system and then I 
procede to the interactive version. 
1. TIlE TGT-SYSTEM IN GENERAL 
The TGT-System of Friedman grew out of the necessity to Verify 
or control a formal grammar. It becomes increasingly di~cult to con- 
trol any formal system beyond a certain size: if one wishes to follow 
the interaction of two or three abstract rules with all their implications, 
The new interactive version of the System has actually been installed at the C.N.U.C.E. 
in order to enable the participants of the Conference to see the systems as it works. I 
take this opportunity to thank the organizers of the conference, the C.N.U.C.E., and 
particularly Professor Faedo, Torrigiani, and Zampolli, once again, for their generous 
support of the demonstration. I also thank my collegues Mrs. Schirmer, Miss Zoeppritz 
and Mr, Henning, who assisted me to prepare the demonstration. I am especially indebted 
Dr. Picchi, who adopted the interactive version to the local cMs-System. 
104 ISTV.~N ~.~TOat 
he may still use his head; for a dozen rules, he will need paper and 
pencil; and for hundreds of rules, he must have a computer. 
As primary objective Friedman wanted to give a computational 
aid to the transformationally oriented linguists. Her system as it stands 
now can, however, be considered also as an attempt to formalize the 
transformational grammar in the strict mathematical sense as well. 
The basic intention of Friedman was not to argue for a specific type 
of generative grammar but rather to Offer a framework as general as 
possible and let the linguist impose restrictions on his particular gram- 
mar. However, it cmmot be overlooked that the starting point of Fried- 
man is clearly CHOMSKY'S Aspects-model. 
Accordingly, it is easy to learn how to work with the TGT-System 
if yOU are familiar with transformational theory. On the other hand, 
you can use to system "to learn" transformational grammar, as a 
tutorial aid. Since we do not want to discuss either the transformation- 
al grammar directly, nor the purely technical details of Friedman's 
System, please, let me presume familiarity with the basic notions of 
generative grammar and refer for the purely notational conventions 
once again to Friedman's book. 
2. THE FORM OF THE GRAMMAR 
The form of the Grammar is strictly prescribed, but as already 
mentioned, it is very close to current transformationalist notation. 
For the TGT-System a grammar consists of a phrase structure, a lexicon, 
and a transformational part. In the first phase of the processing the gram- 
mar is built up according to the users specifications and in the second, 
subsequent phase one sentence (or more) are constructed according 
to the grammar. Each of these major components is subdivided further 
into smaller units. The structuring of the Grammar is indicated by 
keywords, which must be used in certain positions and are anticipated 
by the System. 
Let me shortly comment on some points of this scheme of grammar. 
3. THE TREATMENT OF THE CONTEXTUAL FEATURES 
Friedman introduced a new type of feature, called contextual, 
which comprises Chomsky's strict subcategorization and selectional 
restrictions; i.e., it is all the same for further processing, whether a 
THE INTERACTIVE VERSION OF THE T.G.T.-SYSTEM 105 
contextual rule involves features, like (a) or just category symbols, 
like (b)in (Fig. 1.). 
Contextual Rule 
(a) HUMSBJ----- ( S/( # (ART) N \] + HUMAN \]°/o_ %)). 
0~) DmOBJ = ( V~ ( % NP )>. 
Lexical Entries 
Key \[ + N .... -- HUMAN f, 
Student I + N .... + HUMAN \], 
Imagine I + V .... + HUMSBJ...I, 
Lexical Insertion 
1. Category N V... 
# 
I 
N I 
\[ + N + HUMAN I 
NP lip # 
1 V 
, ., ,., 
Imagine 
I + V +HUMS13Jl 
2. Category V N .... 
S 
# NP VP 
I I N v 
Hul,atz ~ I + V + HUMSBJ. I 
k~ Student \] "~%~1 F 
I+ tIUMAN\[ 
Fig. 1. Side Effects 
But apart from this simplification, the treatment of these contextual 
features is significantly different from that of CHOMSKV'S in,the Aspects- 
model. The main innovation is the concept of the "side effects ", 
which makes the selectional rules independent of the order of inserting 
lexical items into the derivational tree. 
If the contextual feature refers to a node (or nodes) to which a 
lexical entry has already been attached, (as in (1) on Fig. 1) the program 
checks the compatibility of the item with its environment, just as 
in the Aspects-model. If on the other hand (as in (2), Fig. 1) the node 
referred to in the contextual rule is still empty, the new item is intro- 
duced and the consequences of the contextual features, i.e. the feature 
on which the insertion depends, are projected into the invironment. 
f 
106 ISTV~N B£TO~ 
4. THE REPRESENTATION OF TREES 
Note that the output trees are leaned on the side to simplify print, 
ing. In addition the nodes are numbered for ease of reference. These 
numbers can be used, among others to localize the feature, which be- 
long to a specific node, also with higher, non-terminal nodes. Note 
also that features coming from the lexicon are associated originally 
with the lexical entries. After lexical insertion they are adjoined to 
the immediately dominating category node and not to the actual word 
any more. 
5. THE FORM OF TRANSFORMATIONS 
In comparison with the Phrase Structure Rules the notational 
conventions for transformations are less uniform. The notational un- 
steadiness is largely due to the lack of a strict, mathematically founded 
and universally accepted transformational theory. 
There are two notational styles in use; the more popular of them 
is the MIT-Style. (Fig. 2). 
Verbal Description of Passive: 
1. EXCHANGE SUBJECT AND OBJECT 
2. INSERT THE WORD BY AS LEFT SISTER. OF THE AGENT 
3. MAR.K THE MAIN VERB AS PAST PARTICIPLE 
M. L T. - Notation : 
SD # NP X PASS V NP X) 
1 2 3 4 5 6 7 ~ =----------> 
\[ST\]BY+27 SC 1 6 3 4 + PAR 
MITRE - Notation: 
SD % 1 # 2NP % PASS 3V 4NP %, 
SC (PREP < BY > ) ACHLE 2, 
I + PART I MERGEF 3, 
4 ALESE 2, 
2 ARISE 3. 
Abbreviations 
ALESE ... add left sister 
ARISE ... add right sister 
MERGEF... merge feature 
ACHLE ... add by Chomsky ad- 
junction to the left 
Fig. 2. Writing Transformations 
THE INTERACTIVE VERSION OF THE T.G.T.-SYSTEM 107 
This convention for transformation is generally advocated in stan- 
dard introductory works. Accordingly, transformations are written in 
the form of pseudo-rewriting rules, where apparently, the structural 
description (SD) part should be replaced by the structural change 
(SC) part. With other words: you define the input arid you define 
the output. The convention is self explanatory, but perhaps somewhat 
vague. The MIT notation is regarded even by its own adherents 
rather as a convenient short-hand for indicating structural cllanges 
and not as a proper, full scale formalism. 
The other style is the MITRE-notation, which is less known and 
resembles computer commands. This convention defines the !nput 
into a transformation and lists the elementary operations, to be carried 
out on the input tree. The elementary operations shotild be defined 
in advance. On the whole this way of representing transformations is 
more abstract but it can be formalized more readily. Friedman uses 
this style of notation: there is no problem to reformulate a trans- 
formation from the pseudo-rewriting style into the operational 
representation. 
6. '~ THE TRAFHC RULES " 
The purpose of the control program (cv) is to determine m which 
order, and at which point in a derivational tree, a transformation should 
be applied. 
By means of a fORTRAN-like control language (by the so called 
" traffic rules "), the linguist can execute the transformations cyclically, 
i.e. applying the same set of transformations to every clause, he can 
determine in which order the clauses of a sentence should be processed, 
he may change the order of execution depending on certain condition, 
e.g. on the success of preceding transformations etc. This control part 
of Friedman's System provides an enormous generative power, the 
possibilities of which have hardly been discussed in the linguistics. You 
can easily define several successive transformational cycles by the cv 
of Friedman, you can solve the ordering problem of transformations 
by defining unique jumps in order to leave out the execution of a transfor- 
mation, which in a "simple ", cyclically ordered grammar would be 
impossible. 
~v 
t 
108 XSTV~N ~XXOaI 
7. USING THE SENTENCE GENERATOR 
The actual testing of the Grammar is done by the Sentence Gen- 
erator. As already said the Grammar is laid on in the first phase of 
the processing and subsequendy the system should be instructed to 
generate sentences according to the given grammar. Trivially in as 
much as the system generates correct sentences, the grammar is veri- 
fied to the extent the generated sentences are false, the grammar is 
wrong and has to be corrected. 
The sentence generator as such can operate in one of three nodes 
(Fig. 3): 
1. Random Sentences: 
S 
A random sentence will be generated 
2. Predefined Structures: 
S # 
NP ART THE 
N BOY 
VP AUX MOD Q 
TNS PRS 
V PASS 
V READ 
NP PRON WHAT 
# 
The structure will be operated on according to your grammar 
3. Directed Random Generation: 
S RES 
NDOM 
Q 
A random sentence will be generated with the restriction that it will not dominate 
a Q-node 
Fig. 3. Types of Input into the Sentence Generator 
1. It can generate sentencescompletely at random, where a ran- 
dom number generator mechanism controls the selection of grammat- 
ical rules and lexical insertion. All you have to do is to enter the sen- 
tence symbols S. 
THE INTERACTIVE VERSION OF THE T.G.T.-SYSTEM 109 
2. You can predefme a sentence entirely at the level of deep 
structure and let the system check the tree and carry out the transfor- 
mations leading up to .the surface structure. 
3. You can use partially defined input, e.g. defining just the struc- 
ture, but leaving open the lexical insertions, or just specifying a partic- 
ular structural configuration you are interested in, while letting the 
system fill up the rest at random. 
For practical testing the second and the third way of using the 
sentence generator is clearly preferable. The random generator may 
produce spectacular sentences, but practically never the ones which have 
bearing on the problem you are interested in. The sentences delivered 
.by the random generator may be and are revealing, and nobody exper- 
imenting with the system would withstand the temptation to see 
what his grammar would produce "left entirely alone ", but it is 
not suitable for systematic work. You may correct a mistake detected 
by the random generator, but you better test the correction by a pre- 
determined skeleton, otherwise you may get a totally different sentence, 
from which you cannot see whether the error has really been corrected 
or not. 
According to our experience, entirely predetermined structures in- 
cluding lexical entries are the best to test a grammar. In this case you 
can anticipate a normal sentence as the final output of the generator, 
and can immediately decide whether the generation is correct or not. 
There are two input formats: a free, bracketed (FTRIN) format, 
and a fixed tree format (TVJN). It is perhaps a matter of personal taste, 
yet for us the rTRIN, that is the bracketed input, seemed to be more 
convenient. (Fig. 4) 
FTRIN Format: 
S < # NP ( PRON <JEMAND)) VP (NP(AtkT(D) N(BUCH)> 
V ( LES >> MOD < V (HAB> TNS( PtLS >) # ). 
TRIN Format: 
# 
NP 
VP 
MOD 
# 
PRON JEMAND 
NP ART D 
N BUCH 
V LES 
V HAB 
TNS PR.S 
Fig. 4. 
110 ISTVAN B/kTORI 
Usually, the interaction of the phrase structure rules is fairly straight- 
forward, while that of the transformational rules is much more intricate. 
Therefore you can easily predefme a skeleton by using your own phrase 
structure rules "manually " and then let the system apply the transfor- 
mations to the prefabricated input. If you use partially predetermined 
trees, you may be distracted by mistakes, which occur at places which 
are of no interest to you. Note that you cannot correct all errors, at 
least not at once, and therefore you had better concentrate on a few 
points, otherwise you loose sight of you own grammar. 
8. THE OUTPUT OF THE BATCH VERSION 
The original batch output of the ToT-System has been designed 
to provide all possible information about the processing, which the 
linguist may possibly need. First the input grammar is listed, followed 
by the content of the major internal tables, according to which the 
subsequent generation procedes. Then, the process of sentence genera- 
tion is reported in such a manner, that the linguist can follow the sig- 
nificant steps of the processing (Fig. 5 (1)). 
9. TIIE INTERACTIVE VERSION OF TH.E TGT-SYSTEM 
The present interactive version has been developed according to 
the experiences gained by working with the original batch version. 
We have noticed in general that we are interested in the linguistic aspects 
of the derivation, such as changes in the tree, or in the final output, 
but not the actual computation. 
The demand for a more condensed output will be even more im- 
perative in a terminal environment where the time and the output 
should be restricted to a minimum. Therefore we defined a new ad- 
ditional output file, containing just the essential information in which 
a linguist is interested (Fig. 5). 
The original batch protocol enables you to follow the actual flow 
of computation, e.g. in the case of a transformation you get the modules 
called to perform the successive steps of the processing. The interplay 
of the different subroutines is, however, always the same: ANTES T 
calls PASSIV, PASSIV calls ELEMOP etc. Since Friedman's System 
works practically free of error, there is no need to check the subroutine 
THE INTERACTIVE VERSION OF THE T.G.T.-SYSTEM 111 
The Transformations as formulated in the Grammar: 
TRANS 1 PASSIV " PASSIVBILDUNG "l OB. 
SD#INP 2NP 3V 4Vl +PASS 1%. 
SC (PREP I + DAT\[ (VON)) AFIDE 1, 
\[ + PART \[ MERGEF 3, 
2 ALESE 1. 
The Derivational Tree to be Manipulated on: 
1S 2# 
3 NP 4 ART 5 D 
6 N 7 STUDENT3 
8 VP 9 NP 10 PR.ON 
12 V 13 LES 
14 MOD 15 V 16 WEILD 
17 TNS 18 PR.S 
19 Q 20 # 
11 WAS 
The Report on the Successful Completion: 
1. The ordinal batch protocol 
ANTEST CALLED FOR. 2" PASSIV "(AC ) ,SD = 7. RESTRICTION = O. 
TOP= I:S 
ANTEST RETURNS ** 1 ** 
CHANGE. CALL ELEMOP FOR. AFIDE 21 3 
CHANGE. CALL ELEMOP FOR. ALESE 9 3 
2. The new interactive protocol 
PASSIV AFIDE 21 3 ADDED SUBTREE: 21 IS HEADED BY PREP 
PASSIV MER.GEF 12 + PART 
PASSIV ALESE 9 3 
Fig. 5. The Protocol of the Transformations 
calls every time. This information, therefore, can be dispensed with 
for the most purposes. 
We have designed a slightly different, more comprehensive format, 
which contains only the linguistically relevant information. The new 
output format of the interactive version makes a clear reference to the 
input grammar, such as the name of the transformation, the name 
of the elementary operations, the nodes affected by them. In one point 
the interactive version provides information, which has not been explic- 
itly reported in the original batch version. You can follow now also 
the feature operations in the same form as you follow the tree opera- 
tions: the interactive protocol delivers the features names and the actual 
feature value. For a linguist testing feature operations this is an inno- 
112 ISTVAN BATORI 
vation over the original batch version, which suffices to give a hint 
at this point, that the feature operation has been successfully comple- 
ted without further details. 
It should be noted, that batch-output and terminal output are not 
mutually exclusive, the terminal output is a summary extracted from 
the original and placed on a separate file output. The original output 
is, however, still available. The file on which is written is normally 
set dummy, but it can be reactivated and listed, in the very same form 
as in the original version. 1 
10. THE COMMANDS OF THE INTERACTIVE VERSION 
The interactive version on the whole uses a fairly straightforward 
language. The answer to most of the questions is either yes or no (or 
just the first letter of these words). Every answer is prompted; and 
should be answered by saying yes or no. In such cases where an other 
answer is expected the book of Friedman should be consulted. Note 
that in case you want to enter the input skeleton not from the terminal 
you must have the file allocated prior to calling the TgT-System. 
~. THE CONTROL OF THE INPUT 
Summarizing: if you want to run the ToT-System you have to 
define and enter a grammar, give a command for the sentence gener- 
ator, and you have to deliver a skeleton to be expanded (Fig. 6). Orig- 
inally all these three kinds of input were entered in sequence into the 
system on the same file as data. 
It should be noted that the grammar is a part of the input data, 
which is entered and processed in each run. This homogenous input 
is then interpreted by the system as grammar or as input into the sen- 
tence generator according to the internal logics of the program. In 
order to achieve greater flexibility while testing a grammar, we sepa- 
rated the three logically different input into three logically different 
files. The input grammar, usually a text of several hundreds of lines, 
1 The flies 8 and 9, containing system messages have been, however, dropped; they 
were of no interest to ordinary users. 
r 
THE INTERACTIVE VERSION OF THE T.G.T.-SYSTEM 
Bat& Input 
3. Input skeletons'-~ 
2. Generator Commands.-"-~ A 
1. Gram 
Interaaive Input 
I. Gralllllhqr 
2. Generator Commands 
or 
3. Input Skeletons 
Fig. 6. The Reorganization of Input 
113 
) 
is normally already stored on an external device and entered accord- 
ingly. The generator command (the $MatN-card) may be attached 
to the grammar, if not, it is prompted and you may enter it from the 
terminal. 
Similarly, you may predefine input skeletons to be tested and enter 
them just as you enter the grammar as a separate file. You have, how- 
ever, the choice to enter skeletons directly from the terminal. In case 
of interest you may enter as many skeleton as you like. The random 
generator then provides for variation. 
Technically, the separation of the three logically different kinds of 
input has been accomplished by introducing a file variable, which is 
set first to accept the grammar from a permanent data set and then 
changed over to the terminal or an other permanent input data set ac- 
cording to user specifications at session time. 
114 ISTVAN B~TORI 
12. TIlE TREATMENT OF THE ERROR I~iESSAGES 
The same file variable technique is used to control the error mes- 
sages. The error file is set either to the terminal or to the batch file al- 
ternatively. There would be no problem to assign the error messages 
permanently, yet an eventual change of the file requirements in ter- 
minal environment would mean a revision of several hundreds of er- 
ror messages, while a file variable can be controlled by a single instruc- 
tion. 
There is a further problem to be faced and that is the reference 
point of the error message. In the original batch version the error mes- 
sage precedes the actual erroneous line in the grammar or inserted in 
the protocol at the appropriate point. 
In the first case the interastive version does not display the original 
input grammar, and therefore a message that e.g. brackets are opened, 
but not closed or "special character expected ", but not found, and 
the like are not very informative, since the user would be left alone 
to find the critical place in the grammar. Therefore the error messages 
during the processing of the input grammar are preceded by the actual 
line in which the error has occurred. The line numbering will help 
the linguist to localize the erroneous section in the input grammar. 
If on the other hand the error occurs during sentence generation, 
the message will be inserted in the terminal protocol at the appro- 
priate place. 
13. THE CONTROL OF OUTPUT 
Another crucial point is the control of the terminal output. You 
can have the following choices as regards extent of output: 
1) You are not interested in any further details, you ,do not 
want to see the full input tree. In this case you still get: 1., the linear 
representation of the input, 2., the list of transformations which have 
been applied and 3., the output of the transformations, also in the 
linear form. This is the minimal amount of output (Fig. 7): 
2) You wish to see the input tree into the transformational com- 
ponent, you answer to the question PRINTOUT INPUT TREE? 
by saying "yes ". In this case you get also the full output tree of the 
THE INTERACTIVE VERSION OF THE T.G.T.-SYSTEM 115 
ERROR MESSAGES WANTED? 
Y 
INPUT TREE FROM TERMINAL? 
I1 
PRINTOUT INPUT TREE? 
n 
INPUT TO TRANSFORMATIONS: 
# Q NEG WEP. EIN PREIS ERMOEH WERD HAB KOENN WERD PRS # 
~r 
TRANSFORMATIONS WHICH APPLY: 
AUXELIM ALESE 20 19 
AUXELIM ALESE 22 19 
AUXELIM ALESE 24 19 
AUXELIM ALESE 26 19 
AUXELIM ALESE 28 19 
KEIN AFIDE K 13 
KEIN ERASE 0 6 
PASSIV ALESE 12 8 
PASSIV AFIDE 32 8 
logoff in 15 rain 
PASSIV ALESE 8 17 
KNGRUM ERASE 0 28 
VR.ISE ARISE 26 11 
VRISE ARISE 24 11 
VRISE ARISE 22 11 
VRISE ARISE 20 11 
VR.ISE ARISE 17 11 
TOPIC ERASE 0 3 
TOPIC ALESE 8 12 
VERBUM3 ARISE 26 8 
GR.MARK ERASE 0 2 
GtLMARK ERASE 0 30 
INFPAP,. ALADE T 17 
INFPAR SUBSE WORDEN 21 
INFPAR SUBSE SEIN 23 
INFPAR. ALADE EN 24 
MORPHY SUBSE WIRD 27 
INDART SUBSE EIN 14 
PROMOR SUBSE WEM 10 
PUNKT &LADE 41 1 
PUNKT ALADE 43 1 
OUTPUT GENERATED BY TRANSFORMATIONS: 
~r 
,R 
ADDED SUBTREE: 32 
ADDED SUBTR.EE: 41 
ADDED SUBTR.EE: 43 
VON WEM WIRD KEIN PR..EIS ERHOEHT WORDEN SEIN KOENNEN 
QMARK PUNKT 
Fig. 7. 
116 ISTVAN BATORI 
whole generation automatically. You get also the list of the transforma- 
tions which apply displayed; to be more precise you get the list of 
tree operations (Fig. 8): 
ENTER. RANDOM GENERATOR COMMANDS! 
$main trin gen tran. 
INPUT TREE FROM TERMINAL? 
yes 
ENTEP. RANDOM GENERATOR INPUT IN THE FORM OF TI~INI 
+ + + + + 
s 
PRINTOUT INPUT TREE? 
yes 
PRINTOUT FEATURES? 
no 
INPUT TO TRANSFORMATIONS: 
1S 2# 
3NP 
4VP 
5 MOD 
9 PI~ON 15 JEMAND 
7NP 10N 
8 V 16 SEH 
11 TNS 17 PRS 
12 Q 
13 NEG 
# JEMAND HANS SEH PR.S Q NEG # 
TR_&NSFORMATIONS WHICH APPLY: 
NICHT ALESE 18 9 
NICHT ERASE 0 13 
KONGP.UE5 ERASE 0 11 
VERBUM ALESE 8 3 
VERBUM EP.ASE 0 12 
GRLOE ERASE 0 2 
GRLOE SUBSE FtLGZ 6 
OUTPUT GENERATED BY TRANSFORMATIONS: 
1 S 8 V 16 SEH 
3 NP 9 PRON 15 JEMAND 
4VP 7NP 10N 
18 NEG 19 NICHT 
20 FI~GZ 
SEH JEMAND HANS NICHT FRGZ 
1 
ENTER GENERATOR INPUT AS TtLIN 
+ + + + + 
/* 
TRIN .NO MORE INPUTS. 
READY 
14 HANS 
ADDED SUBTREE: 18 
14 HANS 
OR QUIT BY >/~>! 
Fig. 8. 
THE INTERACTIVE VERSION OF THE T.G.T.-SYSTEM 117 
3) You may want to see also the features associated with the 
nodes in the tree - then you respond to the next question of the system 
PRINTOUT FEATURES correspondingly - and you get the features 
displayed both of the input and the output tree. In addition you get 
also the list of transformations applying, now including also the fea- 
ture operations (Fig. 9) : 
call new(tgt 250) 
~r 
WELCOME TO THE INTERACTIVE VERSION OF 
FRIEDMAN\[S TGT-SYSTEM ! 
ERROR MESSAGES WANTED? 
yes 
INPUT TREE FROM TERMINAL? 
no 
PRINTOUT INPUT TREE? 
yes 
PRINTOUT FEATURES? 
yes 
INPUT TO TRANSFORMATIONS: 
NODE 
NODE 
NODE 
NODE 
NODE 
NODE 
1S 2# 
3 NP 4 ART 5 D 
6 N 7 STUDENT3 
8 VP 9 NP 10 PRON 
12 V 13 LES 
14 MOD 15 V 16 WERD 
17 TNS 18 PRS 19 Q 
20 # 
4 ART 
\[ +ART +DEF \[ 
6N 
11 WAS 
"k 
PASSIV A.FIDE 21 3 
PASSIV MERGEF 12 + PART 
PASSIV ALESE 9 3 
NOMIN MERGEF 9 + NOM 
NOMIN MERGEF :10 + NOM 
DATIV MERGEF 3 + DAT 
TRANSFORMATIONS WHICH APPLY: 
ADDED SUBTR.EE: 21 
+ N + MASC -SG + HUMAN + ANIM -PRPNM \[ 
10 PRON 
+ PRON + SG -HUMAN + ANIM + (S/(%_%Q%>> \[ 
12 v 
+ V + EN + STRK + <S/>~NP<(ART)* \[ + HUMAN l>%_b/o>> \[ 
15 V 
I + v + PASS I 
17 TNS 
\[ +TNS +PRS \[ 
#D STUDENTEN WAS LES WERD PRS Q# 
118 ISTVAN BATORI 
KONGR.UE5 MOVEF 10 15 
KONGI~UE5 MOVEF 17 15 
KONGtkUE5 ERASE 0 17 
VERBUM ALESE 15 9 
VERBUM ERASE 0 19 
ERGFIk ARISE 9 2 
INFPAI~ ERASEF 12 SG 
PRS 
INFPAR. AFIDE GE 12 
INFPAk ALADE N 12 
MORPHY SUBSE WIleD 16 
• DEFAR.T SUBSE DEN 5 
DEFAtLT SUBSE DEN 5 
GtLLOE ERASE 0 2 
GRLOE SUBSE FRGZ 20 
OUTPUT GENERATED BY TRANSFORMATIONS: 
-k 
1 S 9 NP 10 PRON 11 WAS 
15 V 25 VCIR.D 
3 NP 21 PREP 22 VON 
4 AtLT 26 DEN 
NODE 
NODE 
NODE 
NODE 
NODE 
NODE 
NODE 
NODE 
1 
READY 
+ SG 
+ PRS 
6 N 7 STUDENT3 
8 VP 12 V 23 GE 
13 LES 
24 EN 
28 FRGZ 
'9NP 
-4- NOM \[ 
10 PRON 
+ PRON + SG -HUMAN + ANIM + NOM + <S1(%-% Q%)) I 
15 V 
+V +SG +PIKS +PASS \[ 
3NP 
+ DAT I 
21 PREP 
+ DAT \[ 
4 APT 
+ AR.T + DEF \[ 
6N 
+ N + MASC -SG + HUMAN + ANIM -PRPNM \[ 
12 V 
+ V + EN + STR.K + PAP.T + <S/<#NP<(AIKT)* I + HUMAN l> 
%-%>> I WAS WIRD VON DEN STUDENTEN GELESEN FP.GZ 
Fig. 9. 
THE INTERACTIVE VERSION OF THE T.G.T.-SYSTEM 119 
4) You may be interested in even more details, for instance in 
some intermediate trees and you have inserted TrACE-cards in the con- 
trol program of the grammar just as they are inserted in the original 
batch version. Now if you answer to the question PRINTOUT IN- 
PUT TREE by saying ALL, you will receive every intermediate tree 
as well, in addition to the input and output tree with features and fea- 
ture operations. Otherwise the TRACt. function returns just the terminal 
string of the derivation. Fig. 10 shows the general logics of the output 
control: 
Printout input tree? 
Printout features? 
Output produced: 
Terminal strings 
Trees 
Features 
Tree operations 
Feature operations 
Intermediate Trees by axse~ 
NO 
+ 
+ 
Yes 
No 
+ 
+ 
+ 
Fig. 10. The Control of Interactive Output 
Yes 
Yes 
+ 
+ 
+ 
+ 
+ 
All 
+ 
+ 
+ 
+ 
+ 
+ 
14. CtOSlNC~F.MARKS 
A grammar developed direcdy with the aid of the TGT-System 
is practically never complete, it generates only a subset of the language 
in question. You may add, change, remove parts of the grammar 
and thus you can easily produce minor variants of the same grammar 
one of which may be preferable over the other. In fact this is the nor- 
mal way to work with the system. 
At the C.N.U.C.E.-installation the/'e was a number of test-gram- 
mars (German, Italian, English and Spanish), offered to the participants 
to try how such testing looks like. The participants of the Conference 
were invited to look at the Grammar Tester as it works. In the Centro 
Nazionale Universitario di Calcolo Elettronico the Transformational 
Grammar Tester was running on a I~M System/360 Model 67 under 
CP-CMS-67. 

REFERENCES 

N. CHOMSK~', Aspects of the Theory of 
Syntax, Cambridge (Mass.), 1965. 

J. FaI~I~lq, T. H. B~DT Lexical Insertion 
in Transformational Grammar, Palo Alto 
(Calif.), 1968. 

J. Fr.I~D~, P,. W. Do~, A Formal 
Syntax for Transformational Grammar, 
Palo Alto (Calif.), 1968. 

J. Fm.~aA~, P. MYst~NsgY, Computer 
Experiments in Transformational Gram- 
mar: The UCLA English Grammar, 
Ann Arbor (Mich.), 1970. 

J. FRIr~MAN, Application of a Computer 
System for Transformational Grammar, 
Preprint No. 14, in International Con- 
~rence on Computational Linguistics, 
Stockholm, 1969. 

J. Fr.mDV_~, Directed Random Generations 
of Sentences, in ,~ Communications of 
the ACM~,, XII (1969) 1. 

j. FRmDV~N, A Computer System for 
Transformational Grammar, in ~ Com- 
munications of the ACM~, XII 
(1969) 1. 

J. Fmr~DMAN, Distribution and Use of the 
Computer System for Transformational 
Grammar, Working paper in Compu- 
tational Linguistics, M-27, The Uni- 
versity of Michigan, 1973. 

J. FRmDmA~, et al., A Computer Model of 
Transformational Grammar (Mathemat- 
kal Linguistics and Automatic Language 
Processing Nr. 9), New York, London, 
Amsterdam, 1971. 

Y. CH. Mom~, Computer Experiments in 
Transformational Grammar: French I, 
Ann Arbor (Mich.), (Natural Language 
Studies Nr. 3), 1969. 

P. I~OSENBAUM, D. I.OCI~AK, The IBM 
Core Grammar of English, Yorktown 
Heights (New York), 1966. 
