ISSA Proceedings 1998 – Argumentational Integrity: A Training Program For Dealing With Unfair Argumentational Contributions
In this contribution we look at the topic of ‘argumentation’ from an ethical perspective. In our research project ‘argumentational integrity in everyday communication’ (funded by the German Research Association since 1988) we are concerned with the conditions, under which people evaluate argumentative speech acts as ‘fair’ or ‘unfair’ as well as with the cognitive, emotional and behavioral reactions to unfair contributions in argumentative discussions. After 10 years of basic research we are now working on a training program based on the results of this research. In our contribution we would like to sketch the main problem dimensions of argumentational (un-)fairness and present the basic concept of the training program.
To start with, let us first illustrate the main problem dimensions by presenting an authentic argumentational episode, which has been recorded and transcribed from a TV-Talkshow. Mr. Krause is a member of the nonsmoker-association, Dr. Troschke is a physician and author. Dr. Troschke and Mr. Krause are discussing, whether smoking is an addiction or not.
Troschke: I try to differentiate the problems in so far, as they can be reasonably discussed. There is a part of smokers who are dependent on the effect of nicotine and who can be labeled as addicted in a very broad sense. This is a relatively small part of smokers who need help to deal with this dependent behavior. The majority of smokers, however, cannot be regarded as addicted, what is simply demonstrable by the fact that, the worldwide most successful method to quit smoking is to decide from one day to the next: I quit smoking.
Krause: For the fifth, tenth, twentieth time!
Troschke: Well, I think, it is extremely difficult to discuss matters on a level where people have different levels of competence and one claims to be able to talk about things one does not know anything about. I do not know, what you really know about addiction problems, about drug addiction or anything else.
Krause: I’m sure, you understand more than I do.
2. Elaboration of problem dimensions
By means of this example we want to illustrate six problem dimensions, which we take up in our training concept.
(1) Is this conversation an argumention? What are the defining characteristics of an argumentative exchange? What is meant by fair or unfair contributions to argumentative discussions?
(2) We assume that participants in argumentative discussions have to consider certain rules of fair argumentation. In our example one of these rules is violated by Dr. Troschke’s contribution: ‘Well, I think it is extremely difficult to discuss matters on a level where people have different levels of competence and one claims to be able to talk about things one does not know anything about.’ We reconstruct this rule violation as a specific form of discrediting of others, that is the denial of competence. Which rules of fair argumentation have to be considered in general, and what type of rule violations have to be distinguished in natural argumentative discussions?(3) If Dr. Troschke really violates one of the rules of fair argumentation, would we reproach him with this violation in any case? Imagine Dr. Troschke was highly emotionally aroused or provoked. Would you still reproach him? How do we come to an adequate evaluation of intentionality and moral blameworthiness? Are there perhaps guilt increasing or guilt decreasing circumstances which have to be taken into account?
(4) Imagine Dr. Troschke indeed violates a rule of fair argumentation. How does the rule violation become manifest at the surface level of the language system? Does Dr. Troschke’s language use tell us, whether he violates the rule intentionally? For example, if we arrive at the conclusion that Dr. Troschke is a highly competent speaker, it appears unlikely that he violates the rule by mistake.
(5) As mentioned before, Dr. Troschke discredits Mr. Krause by his utterance: ‘I do not know what you really know about addiction problems, about drug addiction or anything else.’ Imagine you were the addressee of this utterance: What would have been your reaction? In our example, Mr. Krause chooses an ironical return: ‘I’m sure, you understand more than I do’. With this reaction he addresses the rule violation indirectly. Are there reactions which would have been more effective and adequate?
(6) Leaving the example apart one can raise the fundamental question: Why should contributions to argumentative discussions be fair? This question leads us to the theoretical basis of our training concept which we will present in the following.
3. Overview of the training concept
In our training program, we deal with the problem dimensions just mentioned. The general goal of the training is to raise the awareness of different manifestations of argumentational unfairness and to build up a variety of (fair) reactions to unfair contributions. The training consists of six core units, which are preceded by three introductionary steps and followed by one final step. In the following we will concentrate on the core units, which are elaborated as short-, long- and extension-modules. Theses modules can be combined in a flexible way according to individual needs, expectations and time resources of the participants. The units of our training concept are summarized as:
(I) Warming up
(II) Elaboration of the problem dimensions
(III) Selection of modules
(1) Concept of ‘argumentation’ and conditions of argumentational integrity
(2) Characteristics of argumentational unfairness and standards of argumentational integrity
(3) Blameworthiness and moral evaluation of unfair contributions
(4) Manifestaions of argumentational unfairness
(6) Justification of argumentational integrity
(IV) Feedback and evaluation
In the first training dimension we elaborate in cooperation with the participants a definition of ‘argumentation’ and point out, that the term’ argumentation’ can be used in a descriptive and a prescriptive manner. We consider the prescriptive use of the term ‘argumention’ as the more typical one in everyday communication. In the descriptive usage, an argumentative discussion is conceptualized as a conversation type, which is defined by four characteristis: In an argumentative exchange participants attempt to find a solution to a controversial issue (that is the requirement) by means of a partner-/listener-oriented exchange (that is the process), which is based on reasons for a position and made acceptable to all participants; giving reasons for a position and making it acceptable to all participants are considered to be the goal characteristics of an argumentation. In a prescriptive sense, the reasons should be good reasons and the acceptance should be reached in a cooperative manner. These two goal characteristics, rationality and cooperation, are the basis of an ethical evaluation of contributions in argumentative discussions and can be summarized as ‘generalizability’ (Perelman 1979). In order potentially to reach a rational and cooperative solution to an argumentative discussion contributions should conform to these four conditions wich were formulated on recourse to the literature of argumentational theory (for a comprehensive explication see Groeben, Schreier & Christmann 1993; Schreier, Groeben & Christmann 1995):
Conditions of argumational integrity
(1) formal validity
(3) justice on the content level
(4) procedural justice/communicativity
We have defined the keeping to these conditions as fair, their conscious violation as unfair argumentation (Groeben, Schreier & Christmann 1993). We assume that persons who are engaged in an argumentative exchange know at least intuitively about the prescriptive dimensions as well as the argumentative conditions. That means, when people take part in an argumentative discussion, they implicitely expect, that other participants meet the argumentative conditions. In the training, the participants have the opportunity to explicate and eventually elaborate their implicit expectations by means of various exercises and by means of group discussions and short lectures.
In the second training dimension, the four argumentative conditions are further elaborated. In a first step we specify four charateristics of (un-)fair argumentation, which can be regarded as the ‘negative’ of the four argumentative conditions; on a more concrete level, these characteristics are specified by 11 classes of argumentational rule violations, which have been derived empiricially by a classification of 35 rhetorical strategies representatively choosen from the popular rhetorical literature (Schreier & Groeben 1996; see Appendix). In the present context, it is important to note that it was empirically demonstrated that violations of the standards of integrity are conspicuous and are negatively evaluated both by participants and by neutral observers; that is the standards of argumentational integrity have proven to be empirically valid (e.g. Schreier, Groeben & Blickle 1995; Schreier & Groeben 1996). The goal of this training dimension is to sensitize participants for the 11 standards of fair argumentation and to enable them to correctly identify and label the respective rule violations. For this purpose we have prepared various tasks such as working on written episodes and texts, role playing, group discussion and so on which we will refer to later.
According to our definition of argumentational unfairness as conscious rule violation, the presence of a violation as such may not be sufficient for a personal reproach. Therefore, in the third dimension we focus on the question of blameworthiness and moral evaluation of violations of argumentational integrity. In our basic research, we have conceptualized the unfairness evaluation in analogy to German criminal law which distinguishes between two types of ‘facts’: ‘objective facts’ representing observable features of an action (e.g. to damage a car or to kill another person) and ‘subjective facts’ relating to the actor’s state of awareness in bringing about an offence (e.g. intentional, by negligence, unknowingly). We have transferred this model to the field of argumentative discussions; hence we have conceptualized the evaluation of unfairness as an interplay between the severity of ‘objective facts’ (argumentational rule violations such as ‘distortion of meaning’, hindrance of participation’ etc.) and the degree of subjective awareness in committing a rule violation (Groeben, Schreier & Christmann 1993). The relevance of these two components for the diagnosis and evaluation of argumentational unfairness has been demonstrated in several empirical studies (e.g. Groeben, Nüse & Gauler 1992; Christmann & Groeben 1995; Christmann, Sladek & Groeben 1998). With regard to the example at the beginning of our contribution, it has to be examined, whether the discrediting contribution of Dr. Troschke was committed intentionally or not. As the degree of awareness cannot be observed, but must be inferred, we have specified several indicators of intentionality, which may help to judge the question of intentionality (Christmann, Schreier & Groeben 1996; Schreier 1997); we shall return to this point in the next training dimension. But even considering both, objective and subjective facts, may not be sufficient for a personal reproach. In the same way as in the criminal law -which pronounces a non-guilty verdict if the act is justified – evaluation of argumentational contributions should also consider possible justifications and excuses. A rule violation might be justified if a speaker pursues positive goals by his contribution (e.g. to stir somebody) and it might be excused if the speaker is in a highly aroused state or low in rhetorical and argumentational competence. The empirical results of our research provide evidence that people do, in fact, consider possible guilt decreasing justifications and excuses in their evaluation process, and that an unfairness verdict can be modified in certain cirumstances (Nüse, Groeben, Christmann & Gauler 1993; Christmann, Sladek & Groeben, 1998). In this training dimension, the participants explicate their (rather implicit) structures of evaluation by means of our empirically validated conceptualization of the evaluation process; the explication and elaboration of the respective cognitions may help to substantiate the moral evaluation of unfair contributions in everyday communication and may prevent a rigorous condemnation of any argumentational rule violation.
After having introduced the core training dimensions, we will sketch the other dimensions briefly. In the fourth dimension we will focus on the question of how argumentational rule violations are typically realized in everyday language and how the degree of subjective awareness can be inferred by linguistic indicators on the content level, the interactional level and the argumentational level. In doing so, we refer to the results of our pragmalinguistic analyses, which led to different indicators of these three levels of discourse (Schreier & Groeben 1995). The goal of the fifth training dimension is to strengthen the competence of participants to react adequately to violations of the integrity standards and build up a broad repertoire of different reactions to unfair utterances; we regard appropriate reactions to unfair argumentational contributions as an important protection against manipulation. Referring to the example at the beginning, Krause replies to the Troschke’s discrediting contribution by the following: “I’m sure, you understand more than I do”. This can be interpreted as an ironic aggreement. Apart from this reaction, many other reactions are possible, ranging from cooperative ignoring, through indirect or direct response, up to confrontative, unfair or even breaking-off reactions. Altogether we differentiate between 11 reaction categories, which have been derived from empirical studies (Christmann & Groeben 1993; Schreier, Groeben & Mlynski 1994). In the training we introduce these reaction categories, practice different kinds of reactions by using role-playing techniques and discuss the advantages and disadvantages of the reactions with regard to specific situational circumstances. In the sixth dimension we turn to the question, how the demand on argumentational fairness can be justified and why violations of the argumentational conditions, and of the standards of argumentational integrity, can be criticized from an ethical point of view. By focussing on the justification and legitimacy of argumentational fairness, we sketch appropriate concepts of ‘rationality’ and ‘cooperation’ and we point out how individual bounds of rationality can be overcome by a dialogical principle of rationality. Besides this value-oriented point of view, we discuss possibilites of legitimation under a purpose-oriented perspective and illustrate these possibilities by examples of everyday communication.
4. Instructional design and exemplification
In developing the instructional design we attached great importance in realizing a flexible variety of instructional methods in order to meet the different needs and interests of individual training groups. In accordance with theoretical drafts by Ausubel (1963) and Hermann (1973) we distinguish between four instructional classes:
1. deductive reception learning,
2. inductive reception learning,
3. deductive discovery learning and
4. inductive discovery learning.
We realize these classes by means of a variety of instructional methods (Gudjons, Teske & Winkel 1991) with different ranges (Schulz 1965; 1981; 1996):
(a) instructional concepts (e.g. elements of the Cognitive Apprenticeship approach (Collins, Brown &
Newman 1987), traditional teacher-centered instruction),
(b) settings (learning alone, with a partner, in groups or in plenum) and
(c) instructional actions (e.g. short lecture, role play, video analysis, case method, discussion).
We apply these instructional methods according to different instructional functions (Gage & Berliner 1986). The short modules, for example, generally aim at a condensed impartation of basic and often well structured content. In this case, we see no reason to discard from traditional instructional methods of reception learning, like expert modelling or short lectures given by the training team. For a deeper understanding and for an application and transfer of the acquired knowledge and skills we take recourse to instructional methods of discovery learning like role play, case method or stage-management (realized in particular in long and extension modules). With these methods we also account for principles of modern constructivist approaches, e.g. cooperative and authentic learning, multiple perspectives and multiple contexts (Collins, Brown & Newman 1987; Spiro, Feltovich, Jacobson & Coulson 1991; Gerstenmaier & Mandl 1995). Of course we support the instructional methods by current methods of moderation and media-presentation (Seifert 1993). The whole training method realizes a considerable variety of selectable content and instructional methods. In the following section we give you a brief description of the training dimensions 2 and 5.
In dimension 2 we try to improve participants’ competence with respect to the identification of unfair contributions in argumentative discussions. For this purpose, we treat the 11 standards of argumentational integrity and the corresponding strategies of unfair argumentation. For each of the 11 standards we offer a short, a long and an extension module. For an illustration we present the modules of standard 8 (discrediting of others). A violation of this standard can be found in our introductory example where Dr. Troschke denies Mr. Krause’s competence. Standard 8 has been formulated as follows: “Do not, even by negligence, discredit other participants.” Within this standard we distinguish the following nine strategies of discrediting:
1. argumentum ad personam,
2. ridiculing one’s opponent,
3. denying the opponent’s competence to argue the issue (this is the strategy Dr. Troschke realizes in our example),
4. calling the moral integrity in question,
5. devaluing the opponent’s self-respect,
6. argument by reproaching the opponent with past mistakes,
7. affectation of failure to understand backed by prestige,
8. “psychologizing” and
9. insinuation. At the end of the training, the participants should be able to recognize these strategies in everyday discussions immediately and precisely.
For this purpose we have prepared several instructional versions for the short, long and extension modules. In the short module we provide for two versions. In the first version, the training team gives a compact survey of standard 8 and the respective strategies. The second version has been elaborated in the form of a quiz (see Appendix). First the participants are introduced to an argumentational situation. Then they are asked to generate discrediting utterances which they subsequently have to classify according to the 9 strategies distinguished within the standard ‘discrediting of others’. In the long module, the participants generally consolidate and further elaborate their knowledge, often supported by a practical training of diagnostic competences. In the case of standard 8, the participants analyze and discuss a fictitious argumentative discussion. The extension module generally provides for more complex topics to advance application and transfer. In this case, the participants analyze the video record of an authentic discussion on TV. In all modules of dimension 2 the participants have the opportunity to bring in personal experiences with unfair contributions to argumentative discussions. This will probably lead to questions concerning possible reactions to unfair contributions.
Reactions to argumentative unfairness are the subject of dimension 5. In the following section we give a brief description of this dimension. In dimension 5 we first want to convey a variety of possible reactions to unfair utterances on a cognitive level. In doing so, we will account for interactive, personal and speech characteristics. Subsequently, the participants have the opportunity to practise different reactions in role plays.
In the short module, the participants first generate possible reactions to rule violations. We then present eleven reaction categories, which have been derived from factor analytical studies and replicated in further empirical studies:
1. cooperative ignoring,
2. observant consideration,
3. active clarifying,
4. self protection/defence,
5. defensive ignoring,
6. indirect response,
7. direct response,
8. confrontative discussion,
10. inner breaking off,
11. overt breaking off.
We also explicate the three corresponding dimensions of evaluation:
1. continuing vs. leaving,
2. direct-emotional vs. indirect-controlled,
3. problem-centered/cooperative vs. person-centered/confrontative.
Additionally we discuss criteria for the evaluation of individual reactions. To illustrate these criteria, remember our introductionary example. With his utterance: “I’m sure, you understand more than I do” Mr. Krause realizes an indirect response. This indirect response can be evaluated as a clever reaction for two reasons: On the one hand Mr. Krause indicates that he is not willing to ignore Dr. Troschke’s unfair contribution. On the other hand he shows that he is not interested in risking the discussion at all. In other cases an indirect response might be a too weak reaction, that is, it might not work, e.g. if participants in an argumentative discussion ignore the indirect response and continue to make unfair contributions. In this case a stronger reaction
would be necessary.
The long module provides for role plays to practice the different reaction categories. The participants are enabled to compare and to evaluate the different reaction categories according to the three criteria of evaluation and additionally with regard to the guilt increasing and guilt decreasing circumstances which have been elaborated in dimension 3.
The extension module allows a further discussion of advantages and disadvantages of the reaction category ‘indirect response’ and a discussion of the problem of ‘delayed reactions’.
Our training concept will be evaluated under two aspects: (1) Evaluation of the overall effectivity (‘product effectivity’), and (2) evaluation of single training dimensions during the training process (‘process evaluation’).
To evaluate the overall effectiveness we compare the results of the participants in the training with the results of a group of untrained subjects in a pre-post-design. As the training should improve the diagnosis of unfair contributions and the reactions to unfair contributions, the overall effectiveness is assessed on the level of identification and correct labeling of argumentational rule violations as well as on the level of reactions. For both evaluation levels, we have developed and validated standardized instruments, so we are able to evaluate the overall effectiveness according to criteria demanded in psychological methodology.
The evaluation of single training dimensions is carried out at the end of each dimension by means of standardized exercise-sheets. These exercise-sheets have multiple functions: They serve to recapitulate the main content of the training dimension and give feedback to the participants about their personal success on the respective training dimensions. Last but not least the standardized exercise-sheets serve to evaluate the training dimensions from a process-oriented perspective and give valuable information about the contribution of each training dimension to the overall effectiveness.
At the present stage we are testing and improving the training concept which we have developed. The improved training concept will be carried out in administration- and in business-contexts. We hope that the training concept will help to improve argumentational practice and to find solutions to controversial issues which conform to the goal characteristics of rationality and cooperation.
Standards of Argumentational Integrity
I. Faulty arguments
1. Violation of stringency: Do not intentionally present your arguments in a non-stringent fashion (e.g.: “proof by inconsequent argument”).
2. Refusal of justification: Do not intentionally avoid giving any or intentionally give insufficient reasons in support of your assertions (e.g.: “appeal to mere authority”).
II. Insincere contributions
3. Pretence of truth: Do not make such assertions out to be objectively true which you know to be either false or merely subjective (e.g.: “making false statements”).
4. Shifting of responsibility: Do not intentionally deny, claim, or transfer responsibility to others (persons or institutions) without justification (e.g.: “shifting one’s competence onto someone else”).
5. Pretence of consistency: Do not consciously present any arguments which are not or are only seemingly congruent with what you otherwise do or say (e.g.: “discrepancy between words and actions”).
III. Unjust arguments
6. Distortion of meaning: Do not repeat contributions made by others, your own contributions, or facts in such a way as to intentionally distort their original meaning (e.g.: “changing the meaning of a term during an argument”).
7. Impossibility of compliance: Do not, and be it only by negligence, demand anything of others which you know they will not be able to do (e.g.: “making two mutually exclusive demands”).
8. Discrediting of others: Do not, and be it only by negligence, discredit other participants (e.g.: “denying the opponent’s competence to argue the issue”).
IV. Unjust interactions
9. Expression of hostility: Do not intentionally act towards your adversary in the matter at hand as though he were your personal enemy (e.g.: “attempt to intimidate by being rude”).
10. Hindrance of participation: Do not intentionally interact with others in such a way as to impede their participation (e.g.: “pressuring the others to act”).
11. Breaking off: Do not break off the argumentation without justification (e.g.: “pretending that the issue is really irrelevant”).
Ausubel, D.P. (1963). The psychology of meaningful verbal learning: an introduction to school learning. New York: Grune & Stratton.
Christmann, U. & Groeben, N. (1993). Argumentationsintegrität (XIV): Der Einfluß von Valenz und Sequenzstruktur
argumentativer Unintegrität auf kognitive und emotionale Komponenten von Diagnose- und Bewertungsreaktionen.
Arbeiten aus dem Sonderforschungsbereich 245 ‘Sprache und Situation’, Bericht Nr. 67. Heidelberg/Mannheim.
Christmann, U./Groeben, N. (1995). Diagnosis and evaluation of violations of argumentational integrity: an empirical study. In: F.H. van, Van, R. Grootendorst, A.J. Blair & C.A.Willard (Eds.), Proceedings of the third ISSA conference on argumentation Amsterdam 1994, Vol III (pp. 219-229). Amsterdam: Sic Sat.
Christmann, U./Schreier, M./Groeben, N. (1996). War das Absicht? Indikatoren subjektiver Intentionalitätszustände bei der ethischen Bewertung von Argumentationsbeiträgen. Linguistik und Literaturwissenschaft, 101 (Themenheft Sprache und Subjektivität I), 70-113.
Christmann, U., Sladek, U. & Groeben, N. (1998). Der Einfluß personaler und interaktiver Kontextinformationen auf die Diagnose und Bewertung argumentativer (Un-)Integrität. Sprache und Kognition.
Collins, A., Brown, J.S. & Newmanm, S.E. (1987). Cognitive Apprenticeship: teaching the crafts of reading, writing, and mathematics. In: L.B. Resnick (Ed.), Knowing, learning and instruction. Essays in the honour of Robert Glaser. Hilldale, N.J.: Erlbaum.
Gage, N.L. & Berliner, D.C. (1986). Pädagogische Psychologie. 4th edition. Weinheim: Psychologie Verlags Union.
Gerstenmaier, J. & Mandl, H. (1995). Wissenserwerb unter konstruktivistischer Perspektive. Zeitschrift für Pädagogik 41, 867-887.
Groeben, N./Nüse, R./Gauler, E. (1992). Diagnose argumentativer Unintegrität. Objektive und subjektive Tatbestandsmerkmale bei Werturteilen über argumentative Sprechhandlungen. Zeitschrift für experimentelle und angewandte Psychologie 39 (4), 533-558.
Groeben, N./Schreier, M./Christmann, U. (1993). Fairness beim Argumentieren: Argumentationsintegrität als Wertkonzept einer Ethik der Kommunikation. In: Linguistische Berichte, 147, 355-382.
Hermann, G. (1973). Lernen durch Entdeckung: Eine kritische Erörterung von Forschungsarbeiten. In: H. Neber (Ed.),
Entdeckendes Lernen. Weinheim: Beltz.
Nüse, R., Groeben, N., Christmann, U. & Gauler, E. (1993). Schuldmindernde versus -begründende Zusatzattributionen in moralischen Handlungsbeurteilungen. Gruppendynamik 24 (2), 165-198.
Perelman, Ch. (1979). Logik und Argumentation. Königstein/Ts: Athenäum.
Schreier, M. & Groeben, N. (1995). Unfairness in argumentative discussions: Relevant factors and prototypical manifestations in spoken language. In: F.H. van, Van, R. Grootendorst, A.J. Blair & C.A.Willard (Eds.), Proceedings of the third ISSA conference
on argumentation Amsterdam 1994, Vol III, Amsterdam (pp. 276-286). Amsterdam: Sic Sat.
Schreier, M. (1997). Das Erkennen sprachlicher Täuschung. Münster: Aschendorff.
Schreier, M., Groeben, N. & Mlynski, G. (1994). Argumentationsintegrität (XV): Der Einfluß von Valenz,
Indikatoren subjektiver Tatbestandsmäßigkeit, (Un-)Höflichkeit auf die Diagnose und Thematisierung argumentativer
Unintegrität. Arbeiten aus dem Sonderforschungsbereich 245 ‘Sprache und Situation’, Bericht Nr. 68. Heidelberg/Mannheim.
Schreier, M./Groeben, N. (1996), Ethical guidelines for the conduct in argumentative discussions: an exploratory study. Human Relations 49 (1), 123-132.
Schreier, M./Groeben, N./Blickle, G. (1995), The effects of (un-)fairness and (im-)politeness on the evaluation of argumentative communication. In: Journal of Language and Social Psychology 14 (3), 260-288.
Schreier, M./Groeben, N./Christmann, U. (1995), That’s not fair: Argumentational Integrity as an Ethics of Argumentative Communication. Argumentation, 9, 267-289.
Schulz, W. (1965). Unterricht: Analyse und Planung. In: P. Heimann, G. Otto & W. Schulz (Eds.), Unterricht: Analyse und Planung. Hannover: Schroedel.
Schulz, W. (1981). Unterrichtsplanung. 3rd edition. München: Urban & Schwarzenberg.
Schulz, W. (1996). Anstiftung zum didaktischen Denken: Unterricht, Didaktik, Bildung. Weinheim: Beltz.
Seifert, J.W. (1993). Visualisieren, Präsentieren, Moderieren. 5th edition. Speyer: Gabal.
Spiro, R.J., Feltovich, P.J., Jacobson, M.J. & Coulson, R.L. (1991). Cognitive flexibility, constructivism, and hypertext: random access instruction for advanced knowledge acquisition in ill-structured domains. Educational Technology 31, 24-33.