1Introduction:59 The fundamental problem of mass communication

In most companies and administrations nowadays, written mass communication is made possible by the use of so-called “text modules”. These are frequently used texts or text parts such as salutations, as well as thematic parts such as the introduction to an answer to complaints or information about reimbursement rates in private health insurance. Written, edited and saved electronically only once, they are then combined into whole texts “by hand”. Often, if they fit the communicative purpose, whole sets of text modules are retrieved, for example, in the case of answers to certain complaints.

Mass communication means that a text is intended to be read by many or all members of a group of addressees. Such is the case of general terms of insurance, which are transmitted to everyone who takes out an insurance contract. The text may be revised from time to time, for instance when the legal situation changes. However, it will never address specific aspects of a particular contract. This is impossible due to the very nature of the text, which has been conceived for all potential addressees. The opposite case is that of a text sent to a specific individual on a specific occasion, such as a letter written to a friend on his retirement. This text cannot be sent to another person, or twice to one and the same person.

The core of the problem treated in this chapter is how to strike a happy mean between these extremes, that is, how to draft a text for mass communication in such a way as to make it intelligible for individual addressees. This problem comes close to that of squaring the circle. It cannot be solved by single measures, such as revising some of the texts saved, since these texts are also subject to changes concerning the circle of addressees, the starting points and the relevant facts. Perhaps paradoxically, the solution must be sought in the organization of mass communication, not from a technical, but from a media-related point of view.

2The text-module paradigm and its problems

The process of generating texts by means of text modules can be represented schematically as follows:

As indicated, variants may exist for the beginning (B; e.g., salutation) and the end (E; e.g., complimentary close) of the text. The thematic modules making up the text’s body, whose order will, of course, vary from text to text, have been numbered. Now, where are the problems of this approach, which has been practiced for decades? There are five of them, and only when these are understood in their interrelatedness does the problem of realizing intelligible texts by technical means become apparent in its full magnitude.

This last is the central problem to be solved by communication management. An allembracing solution must be based on prior solutions to problems 1 to 4. However, up to now, no company or institution engaged in mass communication has had any idea, let alone a clear consciousness, of this state of affairs. Most people think that they have learned how to write at school: in reality, today, their ability to write depends not only on their writing competence in the strict sense but also, and equally, on the technological media at their disposal and on their skills in using these. Literacy in the pen-and-paper context is something quite different from literacy in a complex environment dominated by data processing, where advanced computer competence is indispensable. The best piano player will make a fool of himself playing the organ if he has not learned how to use the new instrument’s many stops or, worse, does not even know that they exist. This is what I call the “ignorance problem”. It can only be solved on the basis of information and discussion at the level of decision makers.

3Linguistic communication: A catalogue of misconceptions

Just like laypeople, experts in technical communication harbour many widespread misconceptions about the nature of natural-language, about communication and about the relation between language, thought and action. If we want to solve the intelligibility problem, we must start with these issues, of which the following is a small catalogue.

MISCONCEPTION 1: Linguistic communication follows the scheme of technical communication, that is, sender > message Mi > encoding of Mi > transmission channel > recipient > decoding of Mj.

This model, which goes back to Shannon and Weaver (1948), includes the following “agents”:

Shannon and Weaver’s original model only contained the first three agents and presupposed a message already produced by humans as well as humans capable of understanding the decoded message. The following two theorems hold in this model. First: If Mi has been coded correctly in A, sent and decoded as Mj, then Mi = Mj. Second: C ≠ L, that is, the technical code used for coding and decoding M is not identical to the language in which the message is written. This second theorem, however, does not apply to linguistic communication, where C = L: The thought, or “message”, has no existence independent from the language in which it is formulated. Within certain limits, technical communication is transitive (in the logical sense of the word): if Mi is correctly transmitted by S1 to S2 and decoded as Mj, and then Mj is correctly transmitted by S2 to S3 and decoded as Mk, then Mi = Mj = Mk. This does not generally apply to natural-language communication: If speaker1 says I1 to speaker2, who transmits what he has understood as I2 to speaker3, I2 is generally not identical to I1. Just think of the game Chinese whispers. As a result, natural-language communication falls into a different category from technical communication. Misconception 1 is therefore not just a mistake but a categorisation error, which is why it has been dealt with first here.

MISCONCEPTION 2: Communication is information exchange.

This statement, correct in the technical sphere, does not hold for communication in natural language or writing (for simplicity’s sake, I will omit sign language here). In these media, communication always occurs between human beings, who carry out communicative acts (cf. Austin 1975). This is more obvious in spoken than in written communication. If we ask a friend “Would you be so kind as to bring me a sandwich from the cafeteria?”, she will hardly interpret this utterance as an informationseeking question, but as a request, and so will react accordingly. This is the crucial point since a reaction is always a reaction to an event. The same is true of written communication: On reading a text, readers react to the act of reading, to their interpretation of the text, not to the text itself. The same piece of information (e.g., “We hereby terminate contract XYZ”) will leave one reader cold, while another will break out in a cold sweat. The principle that linguistic communication is always a kind of action applies not only to speech acts such as greetings and requests, but also to written communication. A monthly statement received from our bank is more than just information. It is, in reality, an action whereby the bank fulfils a statutory requirement to inform and summons us to respond: should we fail to respond, we accept that the information is correct. In other words we fully understand the message of this simple communicative act, not when we “decode” it, but only once we have understood how to deal with it and have reacted accordingly. Understanding natural language, therefore, focuses not on what has been said or written, but on the action it implies. And that is what determines how we react to what we have read.

MISCONCEPTION 3: The same thing can be said with different words.

“But that’s not what I wanted to say”, students often reply when you tell them that their answer to a certain exam question was wrong. Students of mathematics would never react this way. They have learned that what counts is only what is on the sheet of paper, the symbolic language of mathematics being unambiguous. But the same does not apply to oral or “normal” written communication. How should the examiner know what the student “wanted to say”? He must rely on what has been written on the paper, word for word. Let us come back to the example given above. The same person who displayed a friendly reaction to the request “Would you be so kind as to bring me a sandwich from the cafeteria?” could be annoyed if we said “Bring me a sandwich from the cafeteria”, and might reply “Ay ay Sir!”, or something similarly sarcastic. As this simple example shows, the sense of what is said or written is always determined by the syntax and the choice of words. We do not first build up a neutral conceptualization that is then translated into natural language. Rather, language is the medium in which we articulate our thoughts, and sometimes we have to write a thought down in order to fully get a grip of it. By explicitly putting what we think before our eyes we come to see which aspects of our thoughts remain unclear.

MISCONCEPTION 4: A medium is simply a means to an end and has no impact on messages transmitted using it.

A means must be sufficient for a certain end; otherwise it would not qualify as a means. Breaststroke is not an adequate means of getting from Hamburg to New York. Sometimes, however, one and the same end can be achieved by different means. This shows that a particular means need not be a necessary condition for achieving a certain end. Things are different as far as media are concerned. The medium of singers is their voice; without using their voice, they cannot sing. The medium is a necessary condition for realizing what is intended. However, not all those who use their voice produce a song. Equally, many of those who have learned to write at school are, nevertheless, unable to write in the sense that they can put a comprehensible thought down on paper. Competence and know-how are needed to use a medium, and since the medium is a necessary condition for expressing what is intended, it cannot be exchanged for another. On the other hand, it is not necessarily enough to achieve the goal. The best voice does not automatically make a singer. In logical terms, means and medium are different notions.

The same conclusion can be reached by considering means and medium from a phenomenological point of view. From this perspective, means are not objects but actions. It is true that we generally use objects as means, for example, a hammer and a nail to fix a picture to the wall. But the hammer and the nail do not by themselves fix the picture to the wall; that is done by the subsequent action of hammering a nail into the wall and hanging the picture on it. Means, therefore, always precede ends in time. That does not apply to media. They are also not objects, but processes, which may also make use of objects. It is not the voice that produces the song, but the use of the voice; it is not the computer that produces the text that we write, but a smoothly running computer program used in a competent way. In this case, the process and the object of the process occur simultaneously, not one after the other. Hence, both from a logical and a phenomenological point of view, means and media are different categories, neither of them objects: means are actions, while media are processes. The consequence is that media, as necessary conditions, are not neutral with respect to what is expressed: the form of the process influences the product. The incomprehensibility of texts in mass communication, therefore, is not just the consequence of bad luck or insufficient wording capability. It results, first and foremost, from the fact that those responsible for the process in general do not understand the relationship between the four problems intrinsic in the process of generating texts by means of text modules, which all result from the “media-based” nature of that process.

4What does it mean for a text to be “comprehensible”?

In the medium of natural language, understanding is not identical to the process of decoding information. Understanding is only complete if you understand what the words mean as an action. But what does that mean? Let us first look at an example. A professor writes to a student who has turned to him with a request as follows: “Unfortunately I cannot write a report for you on . . . But I am happy to help you . . . Please approach my colleague . . .”. How will the addressee understand this message? No doubt as a rejection of his request. He may not even take notice of the rest of the message, or interpret the advice to approach the colleague mentioned as a confirmation of rejection. But, one way or another, he will consider the letter as a whole, and if it contains contradictory elements, the chance is slim that he will accept it in its entirety. Let us look at the situation from the reader’s perspective. He can only understand the letter if he reads it in full and establishes the right connections. For him, the words in the letter constitute an action of which he is the object, and the meaning of an action cannot be split. You cannot thank, or hurt, or advise just a little bit. In any case, the reaction of a letter’s recipient will depend on his understanding of its message as a whole. His reaction corresponds to the action or actions carried out in the letter he has read.

Generally, when writing, we carry out not only one, but several actions. Their order in the text is always of paramount importance, determining as it does the ease or difficulty of understanding the essential message. The following example, taken from a letter sent by an insurance company to a customer, may serve to illustrate this point.

In view of your claims record we are unable to renew your policy as it stands. Nevertheless, we would be happy to assist you. We can renew the policy if you agree to the following amendments: . . . As a precautionary measure, we hereby terminate your contract . . . If you accept the amendments proposed above; please let us know by . . . In that case, the termination of your policy would be null and void.

The situation is a delicate one for the insurance company. It does not want to lose the customer, but nor does it wish to renew the contract on the same terms as before. The problem with the letter lies in the order of the actions it contains:

It is likely that the addressee will take the termination of the policy to be the central message and react accordingly. Will he even consider the option of amending the policy? Will he interpret the termination of the existing policy as a means of coercing him into signing a new one? The following order of speech acts could be an alternative way of proceeding:

In this version the proposal to amend the current policy would be the main action carried out by the letter. As a result, the rest of the text would be read on the premise that the insurance company intends to renew the contract, albeit on different conditions. The termination of the policy would take effect, in this case, as a consequence of a decision taken by the addressee, namely, to reject the proposed amendments. From an informational point of view the two arguments are identical, but the messages transmitted would not be, and the addressee would be likely to react differently in the two cases.

In our context, it is important to understand the logic behind the problem as a necessary condition for dealing with it in a rational manner. This logic is simple, but it does not correspond to our habitual way of thinking. It is based on the following approach. The sense (of an utterance, letter, etc.) is always an indivisible whole. The sense of “Hey, you there!” is more than the sum of the meanings of hey, you and there. The addressee and the other people present will only understand this utterance as an attempt to make contact if they interpret the whole utterance in light of the specific context. The sense of an utterance, as we have seen, is indivisible. From a logical point of view, it is a “super-summative object”, something that is more than the sum of its parts. Such super-summative objects are well-known from everyday life: a football team, for example, composed of goal keeper, defenders, midfielders and strikers, will only function as a team if the different parts, individually and collectively, work as a single unit.

5The hierarchy of levels of comprehensibility

A text that is sent to an addressee – like this present one – should always make sense as a whole. In many cases, however, texts disintegrate into single sections that fit together badly or even contradict each other. At the sentence level, grammar comes to our assistance: it obliges us to integrate words and phrases into one whole, which the addressee can understand, at least at the sentence-level, if he speaks the same language. Beyond this level, grammatical patterns cease to be helpful. What counts is the logical coherence of the sequence of sentences, and sometimes paragraphs, of which the text is composed. These various corollaries will now be briefly discussed in turn.

COROLLARY 1: The message has to be arranged in the medium in such a way that its structure becomes clear to the addressee: the logical structure of the text, of the linguistic actions it contains, should be reflected visually in its layout. In more complex cases, the text should contain section headings or similar devices that help the addressee to understand the structure.

COROLLARY 2: The writer of the text must be conscious of which action is the central one and make clear its relationship to other actions carried out in the text, if there are any. Many writers fail in the task of bringing about logical coherence and a clear logical structure, which constitute the main challenges of the writing process.

That is all that needs to be said about comprehensibility at the text level. At that of the sentence, as we have said, grammar comes to our aid. From a linguistic point of view, a thought is always a clause, albeit a simple main clause and its structure depends on the verb. Logically speaking, verbs are relations, with one (e.g., “John is sleeping”), two (e.g., “John is playing tennis”) or more arguments (e.g., “John accused Jim of bullying”). The verb, therefore, always determines the syntactic structure of the sentence. Apart from the main clause, however, there are often subordinate clauses: “What he heard made him angry” or “The house which he bought is very old”, etc. Simple examples like these pose no problem to the addressee, but it is not so simple in the following, real example:

According to the general terms of contract an inability to work has to be notified immediately to us, at the latest on the first day when benefits are paid, that is, immediately after the end of the maternity leave, in order to allow us to assess whether a claim to benefits exists.

Here no less than four facts are packed into a single sentence, in the following order (order A):

What is the logic of the sentence? Apparently it amounts to the following (order B):

This last fact warrants the conclusion that the addressee has no right to benefits since she has contravened the general terms of contract.

The odd thing about complex sentences with many subordinate clauses is that the order of main clause and subordinate clauses is not fixed. Very often, one can even invert the grammatical function of main and subordinate clause. This changes the sense of the whole sentence or text, but many writers are unaware of this, believing that they alone know what the text is about. There can be no doubt that the logical order of the thoughts is easier to understand in order B than in order A. Order A reflects the analysis from the perspective of the insurance company, which focuses on what the general terms of contract prescribe. Order B, by contrast, looks at the same facts from the addressee’s perspective.

Reformulated in a positive way, the apparently simple advice “Avoid overly complex sentences” amounts to the following: as far as possible, stick to main clauses, and pack only one thought into one main clause. However, such a rule requires a considerable degree of “detachment” with respect to the writing process, something hard to achieve for people who are under time pressure, have a poor command of the topic or insufficient practice as writers. In such cases, most of the task of logically analysing information is left to the reader, who will often have difficulty in coping with it. Reading also has to be learned. Hence:

COROLLARY 3: What you want to say to the addressee should be cast into a logical structure, so take these rules to heart: Write short sentences – long ones are a sign of messy thinking – and package all facts that belong together in a single paragraph.

What remains to be discussed is the level of the vocabulary. At least in common understanding, the logical correlate of a word is a concept. But what are concepts, and what are they good for? Logicians say they serve to distinguish different objects from one another. The simplest concepts, therefore, are nouns. Each noun can serve as a name for a certain set or class of objects: tea, income, argument, etc. Among such classes, we can make finer distinctions: Chinese tea vs. Japanese tea, high income vs. low income, forceful argument vs. weak argument. As these examples show, concepts can be represented not only by simple nouns but also by noun phrases, a grammatical distinction that is irrelevant from a logical perspective. The function of a concept is always to demarcate as sharply as possible one object or class of objects from the rest, and this demarcation is done with words. If we want to explain a concept to somebody else, we will only be successful if we use words whose meanings are clear to everybody. From this fact we can derive

COROLLARY 4: Say what you say with words comprehensible to your addressee. Use technical terms only when necessary to avoid misunderstandings, and always explain their meanings.

From the above, we obtain the following hierarchy of comprehensibility:

The positions in this hierarchy mirror the importance of the three levels for the comprehensibility of a text: it is easier to deduce something more specific from the general context than vice versa. Thereby we arrive at:

COROLLARY 5: First draft a plan and then start writing. Planning is a top-down process, writing is bottom-up. One must have a clear understanding of where one wishes to get to. Experienced writers know that this maxim does not cost, but actually saves time.

6The goal of professional writing: convincing the addressee

What is the goal of a text that we address to somebody? A poem such as Goethe’s “Ein Gleiches” [Another One]60 has no specific addressee, no goal that the writer wanted to attain with this text. Its poetic function is to exemplify what constitutes its beauty or incomparability: calm, harmony, composure, a touch of resignation. . . It does not symbolize, nor represent anything. It simply is as it is, similar in its simplicity to a painting by Mondrian. In contrast, an offer or a reminder serves to pursue a concrete goal with respect to the addressee. In order to clarify this aspect let us go back to what we said about Misconception 2: Communication is never just an exchange of information; it is always an action, which can either succeed or fail. In fact, even the exchange of information is an action that may succeed or fail: If A inadvertently sends certain information to B instead of C, the effects on both B and C will be different from those that would have resulted had the information been sent to the intended recipient. Even saying “Good morning” is an action: If we forget to greet a person known to us, it may have consequences. This is the place to mention another popular misconception:

MISCONCEPTION 5: People act on the basis of information.

As already shown by Aristotle in his Rhetoric, people do not act on the basis of knowledge, but of convictions: that this is or is not the case, that this is or is not necessary, etc. Knowledge is neither a necessary nor a sufficient condition for action. Sometimes we have to act without the relevant knowledge, and sometimes what we take for knowledge turns out to be wrong. As a consequence, the goal of (professional) oral and written communication must always be to convince the addressee. Knowledge and information can be helpful in this endeavour, especially knowledge about the addressee’s convictions: the probability of convincing somebody is undoubtedly higher if we can use their own convictions as premises of our argument. If, on the other hand, we use premises that run counter to those convictions, our chances will be decidedly lower.

7Linguistic communication in complex social systems

8Communication management: What for?

Against the background of what has been stated above, the answer to this question is obvious: because overcoming institutional resistance is a long-term process that can only be carried out, if at all, by the institution itself using the relevant means at its disposal. Communication management, therefore, has to be regarded as a strategic management task which also includes decisions concerning technological media. Focusing on the technological aspects of a medium distracts from its essential functions. By stringing together paragraphs you create a technological, not a logical structure. However, the function of the text is to convince, and, as shown above, this can only be achieved by schemas of argumentation. Last but not least, the use of text modules generally is not accompanied by the philological know-how that is necessary to adequately categorize large amounts of text according to content-related principles.

So what are the tasks of communication management? It has to integrate the (1) linguistic, (2) organizational/institutional and (3) technological requirements for successful communication, where (2) and (3) must be considered as functionally dependent on (1). This can only be done on the basis of a decision matrix that has been discussed, coordinated and decided within the company and is then continually updated. Furthermore, all these activities have to be taken into account in financial planning. Effective communication management is impossible without formalized procedures, at least in bigger companies. And if anyone doubts whether all this is worth the trouble, recall the simple thought that served as a starting point for our reflections: Without successful internal and external communication, there can be no business.

9Can the automation-comprehensibility dilemma be solved technologically?

The typewriter and, before it, the stylus of a Sumerian scribe were up-to-date technological solutions in their time. The success of a solution using present-day technology depends on three necessary conditions.

CONDITION A: The text-module system must be replaced by a procedure able to generate whole texts and variants derived from such texts.

The problem of adapting texts to individual addressees is solved in such a procedure by adding variants to the relevant positions in well-structured documents. The following two examples may suffice to give an idea of how this “whole-text paradigm” works:

Text1 = introduction + body + conclusion

Introduction = (acknowledgement of receipt) + topici

Conclusion = conventional close + salutation

Text2 = introduction + body + conclusion

Introduction = (acknowledgement of receipt) + topicj

Conclusion = conventional close + salutation

Each part of the text has been conceived as part of a whole. For similar topics and situations, text parts can be adapted as apporpriate This guarantees the text’s logical coherence, and thus solves the “reference problem” and the “super-summativity problem” intrinsic to text-module systems. Here is a real example, with two variants:

[Introduction1:] Many thanks for your letter of . . . You state that in our response to your inquiry we agreed only to reimburse your costs up to a sum not exceeding 2.3 times the standard rate. Please accept our apologies for this mistake.

[Body1:] The limit of 2.3 times the standard rate applies to reimbursement of treatment costs in the absence of special justification. In other words, this is the maximum payment we can guarantee you immediately. However, for particularly demanding interventions, reimbursement can rise to 3.5 times the standard rate provided that your dentist supplies us with a detailed justification. Only when we receive this can we guarantee you a reimbursement at more than 2.3 times the standard rate. Your dentist is aware of this and will therefore only charge you such a sum if it is justified.

[Conclusion1:] We hope that this information has eased any concerns you may have had.

Yours sincerely

[Introduction2:] Many thanks for your letter of . . . You state that in our response to your inquiry we did not agree to reimburse the full costs of your treatment. Unfortunately, this is not possible in advance, for the following reason.

[Body2:] In the absence of special justification, we can only guarantee in advance to reimburse your costs at up to 2.3 times the standard rate. Higher reimbursements for particularly demanding interventions at up to 3.5 times the standard rate cannot be guaranteed until your dentist has provided a detailed justification. Only costs that exceed this level cannot be reimbursed.

[Conclusion2:] If you have further questions, we will be glad to answer them. You can reach us on the following phone number: . . .

Yours sincerely

CONDITION B: Media neutrality must be achieved by using a markup language like XML. This solves the problem of “hard-coding”.

More than a decade after the introduction of XML, media neutrality in the organization of technical communication should be standard procedure. Yet in many places this is still not the case. Provided the writing process is carried out in XML, a set of variants like the ones in the example above can be continually adapted to new circumstances without interfering in the technological processes, because in a markup language content and format are strictly separated.

CONDITION C: The process of text generation must be embedded in a systematic company library ordered according to areas of specialization.

This third requirement poses no problems either, since the technological prerequisites, again in the form of XML, are universally available. Embedding the process of text generation in a company library solves the “documentation problem”, and conditions B and C together solve the “technology-dependency problem” of the text-module paradigm.

To sum up, the “management problem” can be solved. But, as we know, all is well and good in theory. The all-important thing is practice.

References

Austin, John L. 1975. How to do things with words. 2nd edn. Oxford: Peregrine Books.

Savigny, Eike von. 1988−89. Wittgensteins “philosophische Untersuchungen”: Ein Kommentar für Leser. 2 Vol. Frankfurt/M.: Klostermann.

Shannon, Claude E. & Warren Weaver. 1949. The mathematical theory of communication. Urbana, IL: University of Illinois Press.

Stetter, Christian. 1999. Schrift und Sprache. Frankfurt/M.: Suhrkamp.

Wittgenstein, Ludwig. 2001 [1953]. Philosophical investigations. London: Blackwell.