Book Reviews 
Linguistic Method 
Essays in Honor of Herbert Penzl 
Irmengard Rauch and Gerald F. Carr, Editors, 
Mouton Publishers, The Hague, 1979, 
628 pp., DM 130, ISBN 90-279-7767-4. 
{Dist: Walter de Gruyter & Co., New York) 
A collection of essays on historical method, the last 
written in Latin with Ciceronian cadences, may not 
seem appropriate for review in a journal devoted to 
computational linguistics. Yet several of the essays 
deal with methods pertinent for synchronic analysis, as 
those noted here may indicate. 
The editors have arranged the essays under four 
topics: I. Linguistic Concepts (pp. 27-160), II. Lin- 
guistic Stratums (161-284), III. Linguistic Systems 
(285-493) and IV. Linguistic Attitudes (495-615). 
Essays have separate bibliographies. A selected bibli- 
ography of Penzl's publications (11-18), an essay on 
linguistic method by Rauch (19-23) and an index 
(619-630) complete the work. Well-prepared and 
presented, the book as a whole, as well as Rauch's 
essay, testify to the affection for Penzl and to the 
respect for his achievements. 
The first part includes a highly interesting essay by 
James W. Marchand entitled Sagena piscatoris, with 
the sub-title "an essay in medieval lexicography" 
(123-137). The essay begins with a reinterpretation 
of the Latin phrase used as title, proposing in contrast 
with a different interpretation that it meant 'net of the 
fisherman' but also that it was a "patristic common- 
place for 'Church'". Thereupon Marchand discusses 
relationships between the SIGNANS and SIGNATUM 
applying during the medieval period, at the same time 
chiding medieval lexicographers for neglecting such 
meanings on the grounds that they are merely figura- 
tive. He vigorously presents the position that 
"throughout the Middle Ages, particularly in the lan- 
guage of the Church, but penetrating all walks of life 
and every kind of discourse words were thought to 
have two kinds of meaning, literal and spiritual" 
(125). His discussion and examples will interest any- 
one concerned with meaning, such as the use of 
Gideon's fleece for 'virgin Mary' -- also 'an unplowed 
field', whether in Latin, medieval Spanish un prado 
bien sin¢ido or medieval German anger ungebrachot. 
Marchand asks that medieval dictionaries include such 
meanings, as well as more remote derivations based on 
medieval etymologies which we consider utterly fanci- 
ful, such as caro datur vermibus, literally 'flesh is given 
to worms' for cadaver 'corpse' and also used when an 
author simply means 'corpse'. Besides such citations 
he provides examples of faulty interpretations by read- 
ers ignorant of medieval conceptions of meaning. 
Apart from the exhortation to medieval lexicographers 
to repair previous inadequacies, Marchand makes the 
point that "in dealing with a passage, we must observe 
not only the immediate linguistic context, but also the 
type of discourse in which it is found" (125). Compu- 
tational linguists have of course propounded such 
views from the earliest days of machine translation, 
when microglossaries were widely advocated, as by 
Reifler, with little impact on general linguistics. It is a 
pleasure to read an article on a somewhat remote area 
of language concern with such illuminating examples 
supporting a more realistic view of language and its 
interpretation. Current texts, especially in technical 
and scientific fields, may not introduce a spiritual 
alongside literal meanings, but terms like bit, hardware, 
program cannot always be interpreted "literally" any 
more than a medieval term like sagena. 
The essay by the dean of Germanic linguists, Hugo 
Moser, also relates to interests of computational lin- 
guists, in pointing to occupational and special lan- 
guages: "Zur Problematik der Fach- und Sonderspra- 
chen" (139-151). Moser discusses their twofold treat- 
ment, on the one hand in sociolinguistic study, on the 
other in pragmalinguistics, in this country known as 
pragmatics. In his discussion, he points out that while 
some occupational languages have been capably stud- 
ied, such as the language of chemistry, many "special 
languages" have not, and he indicates problems that 
must be noted in dealing with these. Like Marchand's, 
his essay also directs attention to complexities that 
must be faced in dealing with lexicon of natural lan- 
guages. Several articles illustrate points he is making, 
e.g., Michael G. Clyne's "Linguistic analysis and the 
1972 Bundestag election campaign" (163-173) and 
Irmengard Rauch's "Semantic naturalness and word- 
building: East German Nur-" (253-263). These illus- 
trate the importance of sociolinguistic and pragmatic 
attention to all linguistic materials, or as Clyne puts it, 
"The main thing is that we do not lose sight of the 
portion of reality we are investigating" (172). 
In the area of grammar Albert L. Lloyd's article is 
of broad interest. Entitled "Prolegomena to a theory 
of Gothic verbal aspect" (327-346), it proposes to 
deal with. aspect generally, at any rate, aspect in the 
Indo-European languages. The article succeeds as well 
as any in this difficult and often debated area. Lloyd's 
definition of aspect agrees with that of other analysts, 
"Aspect refers to the position chosen by the reporter 
as the point of observation on which his report of a 
non-present phenomenon is based" (343). Using 
"normal aspectual pairs in Russian" as a guide, he 
120 American Journal of Computational Linguistics, Volume 7, Number 2, April-June 1981 
Book Reviews Linguistic Method - Essays in Honor of Herbert Penzl 
identifies types of verbs which remain outside the 
area: statals, e.g. 'know', and punctuals, e.g. 'find'. 
The basic problem then concerns what he calls 
"processives." These he analyzes more precisely, 
chiefly with examples from Gothic but also with refer- 
ence to Greek and English, in a treatment too detailed 
for presentation in the space allotted here. Since, 
however, the analysis of verbs in a grammar is of fun- 
damental importance, especially in computational lin- 
guistics and artificial intelligence with their heavy reli- 
ance on case theory, Lloyd's article will repay careful 
study by readers of this journal. His proposed longer 
work following the "prolegomena" may also be kept in 
mind. 
Many other articles not mentioned here are of in- 
terest, also, for specialists in computational linguistics. 
Anyone teaching a course in grammar testing might, 
for example, ask students to program the rules in Jo- 
seph B. Voyles's "The phonology of the Old High 
German Tatian" (441-493). Besides testing the com- 
pleteness and accuracy of his formalizations, a pro- 
grammed phonology would be of no little interest in 
presenting forms in the scantily attested early period 
of German. The availability in English of an article 
"On some linguistic methods" (521-546) onviews of 
Soviet scholarship by a highly regarded scholar, G.S. 
Slur, illustrates a further reason for readers of this 
journal not to disregard this volume. One may dis- 
agree with some of S~ur's statements, for example, his 
"dissatisfaction with typology, whose chief aim is clas- 
sification" (530); actually, typological study of syntax 
has provided some of the best means for insights into 
the development of language and for explanations, 
especially when one moves beyond characteristics like 
adjective:noun and genitive:noun, or preposi- 
tion:postposition structures. Yet this and many other 
articles raise important questions, in a handsome trib- 
ute. In lauding a historical linguist, a highly appropri- 
ate conclusion is an article on views of a major linguist 
concerning the enduring problem of deep and surface 
structure: "Quomodo superficiei grammaticae nexus 
sensusque profundior in Jespersenii Syntaxi analytica 
dicta tractentur" (601-615) by Ladislaus Zgusta. 
Win fred P. Lehmann, University of Texas at Austin 
The FINITE STRING Newsletter 
Announcements 
The Third Annual Conference of the 
Cognitive Science Society 
The Third Annual Conference of the Cognitive 
Science Society will be held August 19-21, 1981, on 
the campus of the University of California, Berkeley, 
in Berkeley, California. In addition to submitted pa- 
pers, there will be four major addresses by speakers 
who are representative of important directions in Cog- 
nitive Science. There will also be four state-of-the-art 
symposia. These will function as tutorial sessions for 
particular subject areas, as well as provide a way of 
learning about the methodologies and concerns that 
the various component fields of Cognitive Science 
bring to bear on common problems. 
Major Addresses 
Robert P. Abelson, Department of Psychology 
Yale University 
Manfred Bierwisch, Central Inst. for Linguistics 
Academy of Sciences (GDR) 
Thomas Kuhn, Department of Philosophy 
Massachusetts Institute of Technology 
William Labov, Department of Linguistics 
University of Pennsylvania 
State-of-the-Art Symposia 
Affect 
George Mandler, Chair, UC-San Diego 
Wendy Lehnert, Yale University 
Katherine Lutz, Harvard University 
Andrew Ortony, University of Illinois 
Amos Tversky, Stanford University 
Cognition and Perception 
Stephen Palmer, Chair, UC-Berkeley 
Geoffrey Hinton, Cambridge 
Paul Kay, UC-Berkeley 
Shimon UIIman, MIT 
Mental Models of Physical Phenomena 
Dedre Gentner, Chair, BBN 
John Seely Brown, Xerox PARC 
Patrick Hayes, University of Rochester 
Philip Johnson- Laird, University of Sussex 
Jill Larkin, Carnegie-Mellon University 
Goals 
Robert Wilensky, Chair, UC-Berkeley 
Richard Fikes, Xerox PARC 
Barbara Hayes-Roth, RAND Corporation 
James Levin, UC-San Diego 
Naomi Quinn, Duke University 
American Journal of Computational Linguistics, Volume 7, Number 2, April-June 1981 121 
