Transactional and Contextual Invalidation between the Parents of Disturbed Families: A Comparative Study^†

The Re-sampling Criteria

After having listened carefully to the 65 available tapes, the author defined objective criteria of exclusion, which were related both to the interviews themselves and to the focus of this study. Thus, all the interviews that were discarded had at least one of four characteristics:

The Classification of the Families Selected

Staff members at the M.R.I. drew largely upon the family's presenting complaints. Those families labeled “disturbed” reported problems in one or more family members. In contrast, no member of the “control” or “non-disturbed” families exhibited symptoms requiring therapy. In 42 of the 46 families reported here, the index patient was an offspring; in 4 cases the index patient was a parent. In 6 families there was more than one symptomatic individual (see Table I). Whenever there was more than one “identified patient,” a family was classified on the basis of the most severely disturbed member. Thus, on the whole the most severe symptomatic disturbance appeared in offspring and was, therefore, assumed to be the hallmark of the family.

Theoretically, if one considers that the family is a system whose patterns stem from contributions of the “sickest” member as well as the “healthiest,” it hardly matters where the pathology is “located.” Most family studies to date, it must be noted, tend to select the family's diagnosis on the basis of the index offspring. Accordingly, this was the criterion for classifying the families at the M.R.I. It is impressive that the present results utilizing the author's approach and those studies selecting the diagnosis of one offspring seem to have yielded congruent results.

The Severity-of-Disturbance Divisions in the Sample

Three different severity-of-disturbance divisions of the total sample were compared here. First, the comparison was between “non-disturbed” and “disturbed.” Second a three-group classification was made: (a) “non-disturbed”; (b) “mildly disturbed” (underachievement, minor maladjustment, enuresis, unspecified psychosomatic complaints, and ulcerative colitis); and (c) “severely disturbed” (delinquency and schizophrenias). Third, in a five-group classification, group 2 was subdivided into “mild family disturbances” and “families containing an ulcerative colitis patient.” Group 3 was composed of families containing a “delinquent” and those with a “schizophrenic” member. The five-group classification was inspired by specific findings on ulcerative colitis (49, 58) and on delinquency (7, 32, 51, 52). In succeeding pages the groups of the five-fold classification will be called groups 1 through 5.

The Limitations of the Sample

The present sample has certain limitations. Qualitatively, there is no matching for age, sex, or chronicity of the disturbances. Moreover, some families have a mixed pathology. Quantitatively, the design is unbalanced. In the dichotomous classification there are 6 control couples and 40 “disturbed” ones. The distribution in the three-fold grouping of “non-disturbed,”“mildly disturbed,” and “severely disturbed” is 6, 18, and 22 couples, respectively. The cases are more evenly distributed in the five-group classification. As Table I indicates, there are 6 “non-disturbed,” 12 “mildly disturbed,” 6 “ulcerative colitis,” 13 “delinquent,” and 9 “schizophrenic” index patients and their families.¹Table I shows the subgrouping within each of the major classes, which is based on the presence of none, one, or more than one kind of disturbance within the index patient.

In spite of the above drawbacks there are at least three strengths that lend validity to this sample. First, sociocultural factors were controlled, thus avoiding the socio-linguistic controversy (39). Second, the diagnoses were carefully made by skilled therapists at M.R.I. Third, the scoring was “blind” and independently done by two well-trained raters.

1. Productivity Scales

A. Number of Lines Spoken: The total number of lines in the type-script were counted for each interview.

B. Percentage of Participation: The total lines spoken by the husband were divided by the total for the couple to obtain percentage scores. The wife's score was the residual of 100 per cent.

2. Unitary Transaction Scales

A transaction was defined here as “the relation of one, usually contiguous message, to another” (46). The interactional methodology hypothesizes a structural relationship between two, usually consecutive, messages by two persons. Hence, the author designated each transaction by bracketing the proper lines in the transcripts so that subsequent ratings would be made on the same segments of interaction. Thus, in the interaction between a husband (H) and a wife (W), the units are counted as follows,

and so on. Each unitary transaction was scored using scales C, D, E, and F.

C. Order-Disorder: This nominal scale (47) classed the information flow. Each transaction has one of five possible qualities: (1) “contiguity,” (2) “partial overlapping,” and (3) “non-disruptive total overlapping,” which are subcategories of order, whereas (4) “interruption” and (5) “disruptive total overlapping” are subcategories of disorder.

D. Interactional Positions: The positions of the interactors can be either (a) “symmetrical,” that is, tending toward the equal status of the participants, or (b) “complementarity,” i.e., differentiating between a “one-up” and a “one-down” party (2, 11, 45, 57).

E. Continuity-Discontinuity: This scale assumes that “every communicational act… may be viewed in some form as an agreement or a disagreement” (23) provided it ensures the “continuity” of the interactional sequence. The continuity categories are: (a) “agreement” (“simple agreement,”“amplification,” or “continuation”), which indicates convergent views; (b) “partial agreement” or “partial disagreement,” which expresses agreement with part of a message and disagreement with the rest of it; and (c) “disagreement” (“negation” or “correction”), which denotes discordant views. These three categories reflect different ways in which the interactors mutually acknowledge each other's messages. On the other hand, the discontinuity category, i.e., (d) “transactional disqualification,” applies when a person fails to label, or mislabels, a “discontinuity” in the interactive sequence. A “discontinuity” is thus introduced by either disregarding the other's message or by uttering a paradoxical one. To quote Sluzki, et al.: “If metacommunicative indicators are absent and the content is incongruent with the context, or if these indicators are present but the content is incongruent with them, there is a disqualification of message a by message b (46). Such transactions suggest avoidance of the decision to agree or disagree with the other. The subtypes of “transactional disqualification” considered are: “evasion,”“sleight-of-hand,”“status disqualification,” and “redundant question.”

Briefly: the continuity-discontinuity scale contains ten subcategories, all of which are mutually exclusive.

F. The Opening Gambit: The first transaction is thought to reflect the extent to which the speaker accepts the task's set. The initial utterance, by either spouse, in response to the interviewer's standard question is scored on a four-point scale: (a) “straight answer” involves directly answering, without any references to the question itself; (b) “labeling and/or repeating the question” involves positive or negative remarks about the question, cues of acknowledgment, and verification of adequate reception of the question (e.g., by echoing it); (c) “asking or giving one's spouse permission to go first” entails claiming or granting the right to speak; and (d) “question response” includes those responses that disconfirm the interviewer's request for an answer (43).

3. The Total Story Scale

G. Relations between the Spouses' Stories: Both spouses' participation was evaluated as contributions to a common product and scored accordingly. The raters drew upon their familiarity with the communicational features used for scoring T.A.T. and Rorschach protocols described by Singer and Wynne (42, 43). Two central features are “closure” (i.e., the joint construction of a meaningful story) and “coordination” (i.e., the ability to share the other's focus of attention and to participate together) between the husband's and the wife's statements. Each rater thus assessed a couple's total interaction, as in a conjoint T.A.T., or Rorschach, in terms of how closely it resembled one of six possible patterns. A numerical score from 1 to 6 was assigned on the basis of the kind and degree of deterioration in “closure” and “coordination”: 1) “concordance” reflects the predominance of different subtypes of “agreement”; 2) “discordance → concordance” refers to those cases in which the spouses begin disagreeing but are able to reach consensus; 3) “discordance” denotes the predominance of the different forms of “disagreement”; 4) “concordance → disconnection” includes those cases in which the spouses begin agreeing but end up with disconnected versions; 5) “disconnection” designates the prevalence of parallel messages; and 6) “contextual blurring” describes those conversations that are marked by lack of “closure” and lack of “coordination” as well. In contextual blurring there is an obliteration of the sense of meaningfulness, tangentializations reccur, and dubious logic is used; the whole interactional context is thus invalidated (43).

The Interpretation of Intragroup Variability

There was high variability within groups 2 (“underachievement,”“minor maladjustment,”“psychosomatic complaints,”“enuresis”) and 5 (“schizophrenias”). The heterogeneity in group 2 supported Stierlin's finding (52) that “underachievement” appears in different kinds of families. In order to identify the underachievers' families that are at high risk for severe disturbances one must look at the specific interaction style between an underachiever's parents.

The variability within group 5 has a twofold explanation. First, it suggests the separation of “childhood schizophrenia” (CS) from “young adult schizophrenia” (YAS). The YAS parents exhibited high “Transactional Disqualification” scores. Contrarily, the CS parents obtained for the most part “Continuity” scores, but their conversations showed at some point a couple of irrelevant speeches or an inappropriate tone that destroyed the integrity of the plot. Hence, these two subgroups shared the end product scored “Contextual Blurring” via different processes. This confirms the differentiation between CS and YAS parents, reported by Singer and Wynne (42).

A second factor of variability in group 5 was the combination of psychological and neurological pathologies in the index patient in one family. This was suggested by the finding of only one significant error in the “Discriminant Function Analysis” of a group 5 couple, who thus appeared similar to the “controls.” This couple had a teenage son who had been successively diagnosed as “mentally retarded” and “schizophrenic.” This enhances Goldfarb's (8) finding that parents who have an offspring with “soft signs” of neurological impairment tend to appear quite intact emotionally. Indeed, such families should be excluded from studies like the present one.

The Differentiation between “Delinquentogenic” and “Schizophrenogenic” Parents

The differentiation between “delinquentogenic” (D) and “schizophrenogenic” (S) parents was suggested mainly by the “Relations between the Spouses' Stories” scale (G). The D parents frequently exhibited “Disconnection,” whereas the S parents showed mostly “Contextual Blurring” in their stories. Additionally, it seemed that the subdivision of “Transactional Disqualification” into “parallel” transactions and “antinomic” ones may help discriminate between the D and S parents at the transactional level as well. The D parents often responded to each other in a “parallel” way, whereas the S parents engaged in paradoxical exchanges. These characteristics appear to be related to Reiss' (35) finding that D parents tend to be “interpersonal-distance-sensitive,” whereas S parents usually bemuddle each other by being excessively “consensus-sensitive.”

The Supersummatory Effects of the Marital Couple

The comparison of “Husband” and “Wife” scores versus the “Couple” scores lends support to the hypothesis that the joint impact of the marital couple is different (“supersummatory”) from the separate effect of each spouse. The analysis of the separate scores showed “Wife” scores to be more uniform than the “Husband” scores within each group and more variable across groups. In contrast, “Husband” scores exhibited a higher intragroup dispersion. This helped explain the overlap of “Husband” scores across groups in some variables that had reached significance at the “Couple” level. Moreover, this finding suggested that the wife's role in marking the couple's style might be more definitive than the husband's. In terms of marital roles one could then speculate that the mother might be the intermediary between the tensions in the marital subsystem and the family.

Another relevant finding was that the separate “Husband” and “Wife” scores suggested a compensatory mechanism in the “mildly disturbed” groups (2 and 3) but not in the “severely disturbed” ones (groups 4 and 5). For instance, “Transactional Disqualification” for husbands and wives differed in group 2. Conversely, those two scores resembled each other in groups 4 and 5. This supports the hypothesis that the least disturbed of two spouses can attenuate the family disturbance (42, 29).

Validation-Invalidation and the “Logical Typing” of Messages

In the early fifties, Bateson introduced Whitehead and Russell's theory of logical types (4) into the study of communication disturbances. Bateson conceived of each message in terms of two distinct aspects, namely “primary communication” and “meta-communication.”“Metacommunication” is of a higher logical type than “primary communication” or the “content” aspect, for “metacommunication” comments (implicitly and/or explicitly) on the communicative process, thereby defining the relationship context. Bateson and his co-workers were intrigued with those incongruencies between logical types that generate a paradoxical relationship pattern, the “double-bind” being its paradigm. In line with Bateson's scheme, it was assumed that “metacommunication” at either the transactional or contextual levels can validate or invalidate a given exchange. Accordingly, the focus here was on two metacommunicative units — i.e., the “singular transaction” and the “total context.”

In the present study, the singular transaction was scored by the “Continuity-Discontinuity” scale (E), whose categories assessed the step-by-step decision-making as the conversation progressed. Each exchange between two partners calls for a decision about the other's message. This decision appears in statements that express decidability, that is, any subcategory of “Agreement,” or “Partial Agreement,” or “Disagreement.” Conversely, the cases of unlabeled or mislabeled undecidability were scored “Transactional Disqualification.”“Transactional Disqualification” should not be regarded as either implicit agreement⁴ or disagreement, since it lacks clear cues of acknowledgment of the previous message. There is no apparent decision about the message. Undecidability, however, does not always nor necessarily result in disqualifications. Therefore, the introduction of an additional category is now suggested for future scoring of those ways of handling undecidability and contextual switches that do preserve the meaningfulness of the relationship context (i.e., mutual validation). The proposed category would permit more clear-cut differentiation between “successful” and “unsuccessful” metacommunication. This study showed the close link between “unsuccessful” metacommunication and “transactional disqualification.”

The total contextual set entails rules between the interactors that can be inferred from the prevalence as well as the absence of certain classes of singular transactions. The emphasis here was on styles of sharing information along three dimensions. The first two, namely, the transmission of meaning (“Closure” versus “Disconnection” or “Contextual Blurring”) and the mutuality of the exchanges (“Coordination” versus “Pseudomutuality”) were assessed by the “Relations between the Spouses' Stories” scale (G). Thus, “Concordance,”“Discordance → Concordance,” and “Discordance” reflect adequate ways of coordinating viewpoints toward a sense of “closure.” Conversely, “Concordance → Disconnection,”“Disconnection,” and “Contextual Blurring” reinforce confusion and “pseudomutuality” (60). Thirdly, the expectations about the interpersonal situation were scored by “The Opening Gambit” scale (F) in terms of the validation (“Straight Answer” or “Labeling and/or Repetition of the Question” or “Asking or Giving Permission to Go First”) or the invalidation (“Question Response”) of the task as a whole.

Validation-Invalidation and “Transactional Thought Disorders”

In the early sixties, Wynne and Singer (61, 62) proposed an “epigenetic and transactional” approach to “thought disorders,” which points to deficiencies in concept formation within affectively toned relationship contexts — i.e., expressiveness among family members. Through the daily interactions with his parents, the child normally develops cognitive skills that allow him or her to interpret interpersonal situations. According to the Wynne-Singer team, then, validation (“mutuality”) would reflect the ability to abstract the relevant properties from either a particular transaction or the total context. This cognitive process presupposes shared focal attention between the inter-actors so that they can in turn monitor and label each other's intentions, feelings, and expectations “internally.” Only thus can any two persons engage in meaningful communication.

Wynne and Singer found that a disordered attentional set (“amorphous” or “fragmented” or “coherent delusional”) (63) toward the listener blurs the meaning of the speaker's messages, thus invalidating the transactions and their context. The extreme forms of “thought disorders” are, in this sense, the schizophrenias. They are characterized by the alteration of symbolic functioning, essentially a breakdown in category formation in the area of social interaction. Yet, this impairment does not necessarily entail the degradation of intellectual potential — e.g., for the sciences and the arts, and instrumental performance in general (1, 15, 24, 25, 26).

To sum up: The emphasis of the Bateson group on “logical typing” of messages, as well as Wynne's and Singer's focus on attentional sets, constitute two relevant ways of conceptualizing the processes by which interactors validate their relationship, or fail to do so.

Empirically, the present study of “transactional” and “contextual” invalidation supports earlier research that has linked parents' communication style to their offspring's psychopathology (4, 7, 9, 10, 19, 20, 21, 22, 23, 27, 29, 32, 35, 37, 42, 49, 50, 51, 52, 54, 58, 61, 62). The author views the systematic focus on metacommunicative dimensions as a most promising line for further inquiry into the structure of family interaction.