Semantics and Syntax in Many an engineer-writer Technical Reports
who thinks
he presents technical information in HERBERT B. MICHAELSON, Metallurgical Laboratories, Sylvania Electric Products, Inc., Bayside, Ν. Υ.
a clear, precise manner is actually indulging in a pleasant self-deception
^?!C5Kf3ffiSSl»K3SM«r--*3.«
. . . Language is a colorful, flexible means of communicating ideas, but this very color and flexibility are at the same time a fertile source of misinterpretation. Although the technical report at its best is carefully worded and rather unemotional, it is never entirely divorced from the in herent defects of language. Moreover, competent engineers frequently are poor writers. If t h e reader of a technical re port on an engineering project is a spe cialist already well acquainted with the subject, loose or obscure wording will not necessarily b e an obstacle to his under standing. But the engineer in a related field who consults the report is likely to have difficulty with word meanings. Care less wording, of course, may confuse readers whether or not they are familiar with t h e subject matter. Some busy engi neers may not even bother to read through a vaguely worded report. T h e choice of words, particularly for the in troductory and concluding sections, should be such that any prospective reader will understand quickly and need not hack his way through a tangled undergrowth of specialized verbiage. This basic problem of intelligibility in technical reporting may b e divided into two phases: ( a ) Semantics, which concerns the com munication of ideas by means of words and t h e maintenance of t h e reader-writer relationship at a mutual level ot under standing. ( b ) Syntax, which concerns t h e accurate expression of ideas by arrangement of words in a sentence according to estab lished usage. Ordinarily we associate the word "se mantics" with the theory of language and the study of the influence of words on human behavior. These academic consid erations are subordinated in this article to the practical aspects of clarity in report writing, and the narrow definition of semantics given in (a) is adopted for the specific case of technical reports. Although the word "syntax" also has a pedanticconnotation, the word as defined in ( b ) has a practical significance to every writer. Poor sentence structure obscures intended meaning. Logical wording, on the other hand, is likely to impress a reader favor ably, and an engineer's technical ability 2416
is sometimes judged solely by the lucidity of his written reports. Semantics and syntax are thus vital aspects of readability in technical reports. Skill and care are needed to build u p word pictures of ideas, even when precise tech nical language is used. Clarity of expres sion is a thinking discipline and a chal lenge to the report writer. Me must suc ceed in making his report understandable to t h e full range of readers who will b e able to use the information, including those not oriented previously to t h e spe cific problem. He should, therefore, show some consideration for readers w h o may have difficulty with his terminology, since every field of engineering has its own special language. This article will at tempt ( 1 ) to show how semantics and syntax are related to the intelligibility of engineering reports and ( 2 ) to offer a few suggestions for making reports under standable. W o r d s and Meanings Language as a means of communication has definite limitations. In all kinds of scientific and engineering literature a graph chart, tabulation of data, photo graph, or mathematical derivation is often worth a thousand words. A technical re port is, however, seldom adequate unless the data and illustrations are accompanied by verbal explanation and evaluation. W h e n we pass from the medium of graphic or mathematical forms to that of word pictures, we step from t h e open
highway of precise representation, w h e r e we can easily recognize concepts by their symbols, into the devious footpaths in the dense forest of language, where w e fre quently "cannot see the woods for the trees." Why, indeed, should words b e an obstacle to understanding? Surely, w e constantly use words to express ideas, and most of us take verbal expression for granted. A clear indication of this problem of word symbolism will b e found on any page of an unabridged dictionary, where a quick glance will reveal that few words have single-valued meaning. Définitions have a peculiar property of shifting with time; some words have special meanings which can b e identified only by their context; others have generalized meanings and might b e interpreted in any of several ways, regardless of context. Misunderstanding occurs when a reader subconsciously selects a definition other than the one intended by the writer. ( T h e expression "the writer" as used throughout this article refers to t h e writer of any technical report. Likewise, "the r e a d e r " means the reader of any technical report. ) T h e fundamental relationship b e t w e e n word, thought, and a physical object is illustrated in Fig. 1. W h e n one reads the word "desk," for example, he immediately acquires by previous association w i t h t h e word a mental image, or reference. He likewise experiences a similar mental impression when he looks at his office desk, or referent. T h e word "desk," of course, has no direct relation to its referent and
Hl and
[KKBERT B. MICHAELSON· is on the inside of things when it comes to semantics syntax in technical reports, for his duties of engineering report editing, report writing, and writing articles for publication for Sylvania Electric Products, Inc., Bayside, Ν. Υ., have given him ample opportunity for first-hand observation. Prior to entering the service in 1943, where he was a second lieutenant and radio officer, Mr. Michaelson was a free lance writer. H i s publications include articles in the Journal of the Franklin Institute, Journal of Applied Physics, Electronics, Proceedings of the Institute of Radio Engi neers, Ceramic Age and other technical and trade magazines. After the w a r he joined the Sylvania Advanced Development Laboratories, Kew Gardens, Ν. Υ., and has recently come to their Bayside laboratory w h e r e he does editorial research. Mr. Michaelson attended Temple University and t h e Polytechnic Institute of Brooklyn, developing interests in electronics, physical metallurgy, and technical ceramics. In addition, he is a m e m b e r of t h e Institute of Radio Engineers.
C H E M I C A L
AND
ENGINEERING
NEWS
cannot be identified with t h e physical object it symbolizes except by the in direct means of t h e reader's mental proc ess. A few r a r e exceptions are onomato poeic words, such as "hiss." T h e bottom line of the triangle in Fig. 1 t h e n becomes a direct relationship because t h e pro nunciation of t h e word symbol is almost identical with the thing it represents. T h e diagram in Fig. 1 represents t h e simplest case of word symbolism, in which one word is related to a single thought and indirectly to a single referent. A more complex symbol may consist of a phrase, clause, or even a sentence. Because of the variety of references a given word or word group can have, information in tech nical reports can be grossly misunder stood. The following example will illus trate the point: A solution of potassium soap is being used to remove t h e surface impurities from the metal parts until a better solution can b e found. By a "better solution" the report writer might mean a better answer to t h e p r o b lem of surface impurity removal, possibly the firing of parts in dry hydrogen. T h e reader of this sentence might easily in terpret a "better solution" to mean an other liquid cleansing agent, such as an acetic acid solution. in addition to the existence of both technical and nontechnical references of a given word, another source of con fusion is the different meanings a word may have to workers in various branches of engineering. For example, an electrical engineer considers a "jack" to be a re ceptacle having electrical connections. T o an automotive engineer, a "jack" is a portable device for lifting vehicles, and to a mechanical engineer employed in t h e textile-machine industry, a "jack" is an element which controls threads in a lace machine. Likewise, the noun "slip" has entirely different meanings to the metal lurgist, the ceramist, the electrical engi neer, a n d t h e civil engineer. Multiple» meanings of this t y p e are not rare; any technical dictionary will yield thousands of such multivalued expressions. This particular semantics problem is not a severe one since each branch of engineer ing has, at least to some extent, an estab lished terminology. More troublesome to the reader are abstract words and phrases. T h e higher the degree of abstraction, the larger is t h e possible n u m b e r of referents. For example, almost any of expressions ( 2 ) to ( 8 ) in the following list might conceiv ably b e used to describe the specific item more accuratelv identified bv expression (1): (1) (2) (ο ) (4) (5 ) (6 ) (7 ) (8)
10 ce. of 3.V IICI at 20° C. a small q u a n t i t y of SN HC1 some 3 N hydrochloric acid some dilute hydrochloric acid a n inorganic acid an acid a liquid a substance
T h e s e expressions are listed in increasing V O L U M E
2 8,
NO.
2 9 » . »
REFERENCE (THOUGHT)
SYMBOL
REFERENT -NOT
(WORD)
DIRECTLY
RELATED-
( T H I N G OR
ACTION)
Fig. 1. Relationships b e t w e e n symbol, reference, a n d referent. ( A d a p t e d from O g d e n a n d Richards) order of abstraction. Obviously, if a writer carelessly chooses one of the high-order abstractions to express his idea, his readers may have to grope for the referent. Worse still, they may misinterpret and be totally unaware of the intended meaning. Loose, indefinite expressions such as "consider able," "many," "very," "somewhat," "slightly," "large," and "small" should b e avoided in any description of quantitative results. To the reader w h o wants specific information, this type of sentence will be inadequate: Considerable time was spent this month in repairing t h e automatic stirring equipment before the investi gation of solubilities could b e con tinued. W h a t is the referent of "considerable time"? A supervisor or contractor, in evaluating the expenditure of research funds, might want to know specifically how much time had b e e n spent on t h e repair work. Another instance is t h e use of the word "conductivity." Unless t h e context completely clarifies the intended meaning, the reader's reference may be either electrical or t h e r m a l conductivity. Also, the "density" of a ceramic body is often used loosely to describe either t h e true density or the a p p a r e n t density, t h e latter being t h e measured value that is uncorrected for pore volume. A final example of multiple meanings a n d misinterpretation is the subject of this very article—"technical reports." T h e reference brought to m i n d by this expres sion will d e p e n d largely on previous ex perience of the individual reader with technical reporting. For t h e engineer en gaged in development projects, t h e refer ence will probably be one of interim progress reports or terminal reports. For t h e product engineer t h e reference m a y b e a routine listing of production a n d quality control data. To t h e sales engi neer, a technical report, may be an analysis of industrial markets; to t h e design engi neer, a description of a new piece of JULY
17,
1950
equipment; to t h e consulting engineer, a detailed problem analysis with recommen dations; to t h e basic researcher, a survey of t h e literature on a given p r o b l e m or a report of a f u n d a m e n t a l laboratory inves tigation. T h e expression "technical re port" used in this article refers to all these various types of reporting, b u t par ticularly to reports w h i c h are not intended for publication. T h e examples cited up t o this j)oint illustrate how the meanings of b o t h tech nical and nontechnical words can b e mis interpreted. This aspect of linguistics should be significant to the report writer. Syntax and M e a n i n g T h e meaning of a single w o r d will fre quently be influenced heavily b y its re lationship to other words in t h e sentence. Chameleon-like, w o r d s take o n color from their surroundings. W h e n a sentence is not p u t together carefully, t h e reader ma, find himself playing a game of Where's the Referent? T h e reader of a technical report will hardly b e in a mood for guess ing games w h e n t h e r e is a stack of other reports on his desk to read, a progress re port to write, a pile of correspondence to take care of, a conference to attend, and pressing work awaiting completion in the laboratory. T h e busy reader will be easily confused by an uncertain relationship b e t w e e n a word and its a n t e c e d e n t . A few examples of ambiguous constructions of this kind will illustrate how the report writer can misrepresent facts unintentionally. First, here is a sentence from a hypothetical report on t h e d e v e l o p m e n t of a n e w sin tered material for use as an electrical heat ing element : T h e next sintered mixture for t h e experimental h e a t i n g unit consisted of 7 7 % talc a n d 2 3 % zinc, b u t this w a s unsatisfactory. T h e antecedent of "this" is uncertain. Is it the m e t h o d of mixing a n d sintering 2417
that was unsatisfactory, or the proportion of zinc, or perhaps the use of zinc itself? Suppose the sentence were modified as follows : The next sintered mixture for the ex perimental heating unit consisted of 77% talc and 23% zinc, but this element was unsatisfactory. At first glance the meaning of this sen tence may appear to be unmistakable— until the possible referents of the word "element" are considered. Which was unsatisfactory» the experimental heating element or the use of the chemical ele ment Zn? A better construction would be: The next sintered mixture, consisting of 77% talc and 23% zinc, could not be used as a heating element becau. c* volume resistivity was too high with this- proportion of talc-zinc. Of course, the antecedent of "this" in the first example about the sintered mix ture might have been clarified by the context. Although the next sentence, paragraph, or section of the report might have explained what was "unsatisfactory," this type of ambiguous sentence structure tends to confuse the reader and to slow up the reading process. But another type of vague wording might not slow him up at all because he may not even be aware of his misinterpretation: All the equipment for this experi mental work has not been received. The last delivery date was July 17. The reader might quickly assume from this that the last time some of the equip ment was delivered was July 17, whereas the writer meant the most recent promise for delivery of the remaining equipment was July 17. Suppose the following construction were used in an attempt to clarify the deliverydate situation: All the equipment for this experi mental work nas not been delivered. The supplier promised to deliver the remaining apparatus on July 17. This might seem to be an improvement —until the significance of July 17 is con sidered. Is it the date of promised deliv ery or the date of the promise itself? A better way of expressing the intended meaning is: July 17 is the most recent delivery date promised for the remainder of the apparatus. Implied meanings are another source of confusion. In the introductory section of a report, for example, this type of con struction is found frequently: For many years no satisfactory meth od was known for sealing together these two materials for use in a vacu um system, but seals have now been made successfully. Where were the good seals first made— as part of this particular development work or in an unrelated recent project? A more common type of ambiguity is a lack of proper subordination of ideas in a 2418
sentence. The skillful writer will ensure clarity of expression by subordinating the dependent clauses and phrases to the main thought. Since this is not done in the following sentence, the meaning is not entirely clear: The carbon brushes in the motor were badly worn and the centrifuge was inoperative. By subordinating the less important ideas in the sentence, the relationship be tween brushes and centrifuge operation is clarified : Since the carbon brushes in the motor were badly worn, the centrifuge was inoperative. The misplaced modifier, too, does its share of deceiving the unsuspecting reader. For example, the placement of the word "only" makes a difference in the meaning of the following two sentences: He only emptied two flasks. He emptied only two flasks. The foregoing examples of faulty syn tax show how much clarity of meaning depends on a logical style of writing; careful diction and precise terminology are in vain unless words are put together to express ideas accurately. Conclusion Because of the intrinsic defects in language itself, there is no foolproof set of rules for achieving clarity in engineer ing writing. Even if all words had singlevalued meaning, the character of a report, or its style, would still depend upon the writer, and faulty syntax could still trip the unwary reader and fling him headlong from the path of understanding. The subtle forms of verbal treachery might be side-stepped if mathematical symbols and equations could be used solely to express ideas in technical reports. Mathematics, however, cannot be utilized without words to express results, opinions, and recom mendations. Although the problems of semantics and syntax in technical reports are too complex for any simple solution, the four suggestions offered here should materially improve the intelligibility of any engineering report: ( 1 ) Include a glossary of technical words and symbols. Although verbal defi nitions are by no means the ultimate solu tion of the semantics problem, the glossary is a most effective device for clarifying word usage. If the definitions are carefully phrased they will prevent misinterpreta tion of special words and symbols used in the report. (2) Use illustrations and tabulations of data wherever possible. Frequently an idea or concept is very difficult to express adequately and concisely with words alone. A photograph, diagram, or list of data often will be useful For supplement ing a brief verbal explanation. (3) Have the rough copy edited by someone else. Faulty syntax is generally difficult for the report writer to detect; he may read the draft many times without recognizing poor sentence structures. A qualified reviewer, however, will quickly find ambiguities in the rough copy which would go unnoticed by the writer. More CHEMICAL
over, if the report is likely to be read byworkers in other fields of engineering, the draft should be reviewed by a competent editor who is not a specialist in the topic of the report. The critique of such a reviewer will usually strike deep to the core of the semantics problem. In addi tion, the writer should have his report edited for technical accuracy by at least one specialist in the writer's field of engi neering. (4) Be concise. Brevity is the soul of clarity. Regardless of these recommendations, the comprehensibility of a technical re port can never be taken completely for granted. The ability of a reader to under stand is based largely on his previous experience, and hardly any piece of writ ing can reach all experiential levels. The meanings of every-day words in some contexts, moreover, are often more ob scure than those of technical expressions. The limitations of language are thus a challenge to the skill of any writer. The author of this article would not dare to claim that all his own technical reports and published articles are completely un derstandable to every reader. But lan guage, regardless of its imperfections, is an indispensable medium of communication. An awareness of the tyranny of words will help the report writer to present informa tion clearly. Selected Bibliography Bennet, J. B., "The Technical Report and Its Problems," C H E M . ENG. NEWS,
26, 2435 (Aug. 16, 1948). Chase, Stuart, The Tyranny of Words, Harcourt, Brace and Co., New York, 1938. Crawford, D. M., "On Engineering Writ ing," Mech. Eng., 67, 607 (1945). Goodell, J. D., "The Written Word," Radio News, 120 (Nov. 1943). Gray, D. E., "Making Technical Reports Understandable," J. Chem. Edu cation, 25, 226 (1948). Gray, D. E., "How Readable is Your En gineering Report?" /. Chem. Edu cation, 26, 374 (1949). Hayakawa, S. I., Language in Action. Harcourt, Brace, and Co., New York, 1941. Howell, A. C., A Handbook of English in Engineering Usage, John Wiley & Sons, Inc., New York, 1940. Korzybski, A. B., Science and Sanity, Sci ence Press Printing Co., Lancaster, Pa., 1941. Lee, I. J., The Language of Wisdom and Folly, Harper and Bros., New York, 1949. Longacre, Α., "A Semantic Approach to the General Physics Laboratory," Am. Jour. Phys., 17, 413 (1949). Michaelson, Η. Β., ^'Techniques of Edi torial Research in Electrical Engi neering," four. Frank. Inst., 247, 245 (1949). Michaelson, Η. Β., "Clarity in Technical Writing," Proc. 1. R. E., 37, 1455 (1949). Miller, J. Α., "Technical Writing," Civ. Eng., 18, 43 (May 1948). Nichols, W. T., "The Technical Report Bugaboo," C H E M . ENG. NEWS, 26,
602 (Mar. 26, 1948). Ogden, C. K., and Richards, Ι. Α., The Meaning of Meaning, Harcourt, Brace and Co., New York, 1938. Walpole, Hugh, Semantics, W. W. Norton and Co., Inc., New York, 1941. AND ENGINEERING
NEWS