the 1982 interpersonal circle: a taxonomy for ... · described corresponding overt interpersonal...

Psychological ReviewVOLUME 90 N U M B E R S JULY 1983

The 1982 Interpersonal Circle: A Taxonomy forComplementarity in Human Transactions

Donald J. KieslerVirginia Commonwealth University

The purpose of this article is to integrate previous theory and research addressinginterpersonal complementarity, a construct that is central to refined and extendedresearch and clinical applications of interpersonal theory. The article first de-scribes the 1982 Interpersonal Circle, which the author constructed as a com-prehensive taxonomy of the domain of two-dimensional interpersonal behavior.The 1982 Circle integrates and expands the content of four major adult inter-personal measures (LaForge & Suczek's Interpersonal Check List, Wiggins's In-terpersonal Adjective Scales, Lorr & McNair's Interpersonal Behavior Inventory,and Kiesler et al.'s Impact Message Inventory) to provide a circle taxonomyconsisting of 16 segments, 128 subclasses, 2 levels, and 350 bipolar items. Second,the article reviews previous conceptions of interpersonal complementarity and,using the 1982 Circle as a theoretical and operational guide, derives 11 propo-sitions of complementarity as they apply in personality, psychopathology, andpsychotherapy.

Various writers argue that interpersonaltheory has significant implications for thestudy of personality, abnormality, and psy-chotherapy (Adams, 1964; Anchin & Kiesler,1982;Bierman, 1969; Carson, 1969; Chance,1966; DeVogue & Beck, 1978; Foa, 1961;Kiesler, 1979, 1982b; Kiesler, Bernstein, &Anchin, 1976; Leary, 1957; McLemore &Benjamin, 1979; Wiggins, 1982). Unfortu-nately, the full power of interpersonal theoryhas yet to be convincingly demonstrated,and, as Wiggins (1982) observes, the inter-personal "approach has not been accordedas prominent a place in the mainstream ofclinical thought as its proponents would de-sire" (p. 183).

One reason for this attenuated impact isthat interpersonal researchers have directedthe bulk of their energies to the area of per-sonality, with considerably less theoretical or

Requests for reprints should be sent to Donald J.Kiesler, Department of Psychology, Virginia Common-wealth University, 806 West Franklin Street, Richmond,Virginia 23284.

empirical attention being devoted to issuesof psychopathology and psychotherapy. An-other reason is that available two-dimen-sional interpersonal measures, although gen-erating large bodies of empirical findings, areneither theoretically nor empirically compre-hensive or precise enough to validly test cen-tral propositions of interpersonal theory orto guide concrete applications to assessmentand therapy for abnormal behavior.

The purpose of this article is to integrateprevious theory and research addressing in-terpersonal complementarity, a construct thatis central to refined and extended concreteapplications of interpersonal theory. I willfirst present the 1982 Interpersonal Circle, ataxonomy that represents my schematizationand integration of adult two-dimensional cir-cle measures developed by previous investi-gators. As I will document, none of theseprevious versions possesses the full range oflogical and empirical properties necessary forprecise and comprehensive derivation oftheoretical propositions and empirical appli-cations. Second, using the 1982 Circle as a

Copyright [983 by the American Psychological Association, Inc.

185

186 DONALD J. KIESLER

theoretical and operational guide, I will turnto a systematic review and derivation of prop-ositions of complementarity that are basic toheuristic applications of interpersonal theory.

The 1982 Interpersonal Circle

Wiggins (1982) provides an excellent sum-mary and critique of two-dimensional andother interpersonal measures. His analysis,critiques, conclusions, and especially his owncircle measure served as fundamental guide-lines for development of the 1982 Circle. Thelatter integrates the content domains of thefollowing four major adult, two-dimensionalmeasures: (a) the Interpersonal Check List(ICL; LaForge & Suczek, 1955; Leary, 1957;LaForge, Note 1), (b) the Interpersonal Be-havior Inventory (IBI; Lorr & McNair, 1965,1967), (c) the Interpersonal Adjective Scales(IAS; Wiggins, 1979, Note 2), and (d) the Im-pact Message Inventory (IMI; Kiesler et al.,1976; Perkins et al., 1979).

Essentially, the 1982 Circle expands thework of Wiggins (1979, 1982) by providinga taxonomy consisting of a large item sampleof the universe of possible two-dimensionalbehaviors. From this item sample, alternativeforms of self-report and rating measures canbe constructed and, from a match of categorycontent, new measures using different itemformats can be derived. The 1982 Circle thusserves as a comprehensive taxonomy of thedomain of two-dimensional interpersonal be-havior to which extant and future measuresor constructs may be compared, contrasted,and/or anchored.

The basic problem with the four majoradult, two-dimensional circle measures isthat none incorporates the full range of thefollowing theoretical, methodological, andempirical features necessary for a compre-hensive and heuristic interpersonal circle.Most of these individual features have beenarticulated clearly by the authors of the pre-vious measures, yet the fact remains thatnone of these empirical translations incor-porates all of them.

1. An interpersonal circle defines a cir-cular array of categories (segments)—factoranalytic findings usually yielding 16 (A-P)—that operationalizes the domain of interper-

sonal behavior. The ordering of categories iscircular in that it is without beginning or end.

2. On the circular continuum, categoriesor segments located at polar ends of circlediameters are defined as classes of interper-sonal behavior representing behavioral con-trasts and/or semantic opposites. Hence,each of the 16 segments should be highlynegatively correlated with its opposite andshow zero correlations with theoretically or-thogonal segments. Wiggins (1982) docu-ments that categories present in circle mea-sures other than his IAS, and especially in theICL, are not consistently arranged so thatlogical opposites appear directly across theeight diameters of the circle. This representsa deadly criticism of any interpersonal mea-sure, because the principle of opposites (Ben-jamin, 1974; Wiggins, 1982) defines a basiccircle characteristic without which it is im-possible to derive precise and consistent pre-dictions of interpersonal complementarity.

3. The circular array represents a two-di-mensional Euclidian space reflecting the jointaction of two basic interpersonal dimensionsor motivations, almost universally designatedControl and Affiliation. These dimensionsdefine, respectively, the vertical and horizon-tal axes of the circle. A large body of researchreviewed by Berzins (1977), Bierman (1969),Carson (1969), DeVogue and Beck (1978),Foa (1961), and Wiggins (1982) convincinglydemonstrates that interpersonal behaviorrepresents the joint expression of these twounderlying dimensions. Hence, this criterionreflects the assumption that pairs of inter-actants, in their daily transactions, are ne-gotiating mutually satisfactory definitions re-garding who is going to be more or less incontrol or dominant and what is to be thecharacteristic level of friendliness or hostility.

4. Each of the 16 segments is a blend ofthe two axis dimensions reflecting mathe-matically weighted combinations of Control(-4 through 0 to +4) and Affiliation (—4through 0 to +4). That is, force fields emittedby human interactants are defined in termsof the mathematical weightings or loadingsof each one's respective behavioral segmentson the two axes of Control and Affiliation.For example, a person's behaviors defined atSegment H exhibit -1 friendly and -3 dom-inant components, whereas a person's be-

INTERPERSONAL CIRCLE 187

haviors at Segment B show —1 friendly, +3dominant elements. This criterion is essentialbecause validation of propositions of com-plementarity requires segment definitionsthat match the precise circle mathematicsspecified originally by LaForge and Suczek(1955) and Leary (1957).

5. Empirical intercorrelations among the16 segments should reveal a circumplex or-dering (Guttman, 1954) wherein segments ofinterpersonal behavior adjacent on the cir-cumference are positively correlated, andsegments opposite on the circle are negativelycorrelated (LaForge & Suczek, 1955). Thisordering is confirmed by the presence of aparticular pattern of coefficients found in thematrix of intercorrelations among the 16 seg-ments. A circumplex pattern exists in a ma-trix when "the highest correlations are nextto the principal diagonal which runs from theupper left to the lower right corner. Along anyrow (or column) the correlations decrease insize as one moves farther away from the maindiagonal and then increase again" (Lorr &McNair, 1965, p. 824).

6. For each of the 16 segments, the radiusof the circle represents the intensity or ex-tremeness of corresponding interpersonal be-haviors. The degree of extremeness or ab-normality of a particular behavior is repre-sented precisely by its distance from themidpoint of the circle (LaForge & Suczek,1955; Leary, 1957).

7. To permit precise assessment of the en-tire continuum of mild to extreme interper-sonal behaviors, a circle should provide atleast two (ideally more) levels of definitionsand operationalizations for each of the 16segments. This criterion asserts that level dif-ferentiation is crucial for validation of inter-personal theory. For example, interpersonaltheorists (e.g., Leary, 1957) define abnormalbehavior in two ways: (a) as rigid adherenceto 1 or a few of the 16 segments of the in-terpersonal circle; and (b) as behaviors at ex-treme levels of 1 or a few circle segments.Differentiation of abnormal from normal in-terpersonal behavior, therefore, requires ameasure sensitive to the entire radius contin-uum (mild to severe) of each circle segment.If interpersonal researchers are to make sig-nificant contributions to the diagnosis ofmaladjusted behavior by using the interper-

sonal circle as an alternative to Diagnosticand Statistical Manual of Mental Disorders(DSM-III; American Psychiatric Association,1980) nosology (Adams, 1964; McLemore& Benjamin, 1979), then the extreme levelsof each of the 16 segments need to be clearlyseparable from milder and more normallevels.

8. To facilitate precise discriminationsamong interpersonal behaviors, definitionsand operationalizations of each level of a par-ticular segment should show minimal se-mantic/behavioral overlap with adjacent orother segments.

9. For comprehensive assessment and pre-cise discrimination of interpersonal behavior,if possible, each segment should be elabo-rated further by defined subclasses of behav-ior at each level.

10. Circle items describing interpersonalbehaviors should be in the form of eitherunambiguous adjectives or verb phrases de-scribing overt behaviors, or both. One of thefrustrations of interpersonal checklists is thesemantic ambiguity inherent in many, if notmost, adjective items. For example, describ-ing a person's behavior as "severe" can de-note either a person who is strict or rigorousin judging and disciplining others (which de-fines his or her behavior at Segment D of thecircle) or a person who behaves harmfully,abusively, ruthlessly, sadistically, and so forth(which defines his or her behavior at SegmentE). In response to this problem Lorr andMcNair (1965) developed their IBI by trans-lating ICL adjective items into items thatdescribed corresponding overt interpersonalbehaviors. Buss and Craik (1981) emphasizedelineation of interpersonal acts in contrastto dispositions reflected by adjectival behav-ioral descriptions. Their act'in-context unitsanchor interpersonal assessment in both overtbehaviors and situations simultaneously.Hence, if a circle taxonomy is to guide reli-able and valid characterizations of interper-sonal behavior, it must be composed of un-ambiguous adjective items and/or items thatdirectly reference overt interpersonal behav-iors.

11. Circle definitions should be suffi-ciently operational in form to permit inde-pendent investigators reliably to cross-classify


items from interpersonal measures at one ofthe two levels of 1 of the 16 circle segments.

12. A circle should permit precise deri-vation of propositions of complementarityfor each of the two levels of each of the 16segments, as well as for circle octants andquadrants. Although each of the four majoradult two-dimensional measures defines 15or 16 segments, none has been able to classifysubjects' behaviors reliably into individualsegments (Wiggins, 1982). Instead, the bestthat seems presently possible is classificationof subjects' behaviors into octants. Also, in-vestigators disagree as to the most heuristictactic for data analysis, with some arguing foranalysis of segments (LaForge, Note 1), somefor octants (Leary, 1957), and others forquadrants (Carson, 1969). What this last cri-terion asserts is that for derivation of preciselaws of complementarity, measures are re-quired that can classify interpersonal behav-iors at one of the two levels of 1 of the 16segments. Precise placement into segmentlevels seems necessary if we are to developthe full power of the circle to assess the entirerange of individual differences in interper-sonal behavior.

Of the four major adult two-dimensionalmeasures (ICL, IBI, IAS, IMI) none incor-porates all 12 of these essential criteria—allfall short on at least 5. In contrast, the 1982Circle taxonomy was designed to meet ex-actly all 12 criteria.

Figure 1 depicts the 16 segments of the1982 Circle. Table 1 lists the bipolar sub-classes that define mild-moderate versus ex-treme levels of each of the 16 segments, aswell as the number of circle items that op-erationalizes each.1

The 1982 Circle items were constructed bythe laborious and obsessive pursuit of an ex-haustive list of synonyms and antonyms foreach item present or implied in the ICL, IBI,IAS, and IMI. When possible, adjectives werealso translated into verb-phrase descriptionsof overt interpersonal behaviors. Because ofthe superior circumplex structure Wiggins(1982) has demonstrated for his measure,IAS items were used as initial markers forsegment location and definition. This was fol-lowed by multiple classifications and reclas-sifications of items as they were arrangedunder each level of each of the 16 segments.

Additional reclassifications occurred as it be-came clear that each segment could be fur-ther differentiated into several subclasses.The final step in this iterative process wasclassification of synonym-antonym, bipolaradjective, and verb-phrase items under eachsubclass of each level of each segment pairuntil the final list seemed relatively exhaus-tive or representative. The major aids I usedin this task, in addition to the items from theprevious measures, were the Random HouseDictionary: Unabridged Edition and Double-day's Roget's Thesaurus in Dictionary Form,with backup from Webster's New CollegiateDictionary and Webster's Collegiate The-saurus.

The final list of 350 bipolar items has anitem format distinct from all previous inter-personal measures. Instead of listing unipolaradjective or verb-phrase descriptions, each ofthe 350 bipolar 1982 items defines an inter-personal behavior for both a particular seg-ment and for its behavioral/semantic oppo-site (located at the segment directly across thediameter of the circle), as illustrated by thefollowing Segment A and Segment I items:

1. able to give orders easily led

4.

talks others intodoing what hewants

often assumesresponsibility

influential

lets others makedecisions

avoidsresponsibility

ineffective withothers

Each of these four items simultaneously de-fines a person's behavior on Segment A andon its opposite, Segment I. Item 1 assessesboth A:Dominant ("able to give orders") andits opposite, I:Submissive ("easily led").

As a result of this bipolar format, one poleof the 350 items defines 8 of the 16 segmentsand the other pole defines the other 8 op-posite segments. If this bipolar format were

1 The complete 1982 Circle taxonomy including seg-ment, level, and subclass definitions as well as their re-spective bipolar items (a total of 350) can be found inKiesler (Note 3). The manuscript, which is availablefrom the author, also classifies each of the items fromthe ICL, IBI, IAS, and IMI under its corresponding 1982Circle segment and level.


Figure 1. The 1982 Interpersonal Circle.

eliminated, each bipolar item could yield twounipolar items: For example, Item 1 couldyield la ("able to give orders," Segment A)and Ib ("easily led," Segment I). However,for the 1982 Circle taxonomy the bipolar for-mat was considered essential for two reasons:(a) to ensure that the principle of opposites(Benjamin, 1974; Wiggins, 1982) was builtinto every aspect of the 1982 Circle, and (b)to eliminate the behavioral/semantic ambi-guity so often inherent in adjective items.

Figure 1 and Table 1 show that each of the16 segments is assigned three separate labels.

One label designates the entire continuum(the circle radius) of interpersonal behav-iors constituting a particular segment(e.g., A:Dominant, E:Hostile, I:Submissive,M:Friendly). A second label designatesthe mild-moderate level of a particularsegment continuum (e.g., A] Control-ling, E]:Antagonistic/Harmful, Ii:Docile,M, :Cooperative/Helpful). A third names theextreme level of a particular segment(e.g., A2:Dictatorial, E2:Rancorous/Sadistic,I2 :Subservient, M2 :Devoted/Indulgent).

(text continues on page 192)


Table 1Levels and Subclasses for the Eight Pairs of Opposing Segments for the 1982 Interpersonal Circle

Segment, level, and subclass Opposing segment, level, and subclass

A:Dominant (38)

A,:Controlling(21)

a. Leading/influencing (7)b. Active/self-assertive (4)c. Strong/managing (5)d. Taking-charge (5)

A2:Dictatorial (17)

a. Commanding (5)b. Overbearing (5)c. Bossy (3)d. Tyrannical (4)

B:Competitive (45)

B,:Critical/Ambitious (25)

a. Energetic (7)b. Enterprising (6)c. Competitive (7)d. Critical (5)

B2:Rivalrous/Disdainful (20)

a. Driven (4)b. Daringly shrewd (4)c. Rivalrous (6)d. Disdainful (6)

CMistrusting (51)

C, :Suspicious/Resentful (28)

a. Vigilant (4)b. Suspicious/jealous (9)c. Cunning (5)d. Resentful (4)e. Covetous/stingy (6)

C2:Paranoid/Vindictive (23)

a. Hypervigilant (4)b. Paranoid/blindly jealous (5)c. Crafty/exploitative (6)d. Vindictive (4)e. Avaricious/envious (4)

I:Submissive (38)

I,:Docile(21)

a. Following/complying (7)b. Passive/acquiescent (4)c. Weak/yielding (5)d. Obedient (5)

I2:Subservient (17)

a. Servile (5)b. Spineless (5)c. Submissive (3)d. Slavish/fawning (4)

J:Deferent (45)

J,:Respectful/Content (25)

a. Insolent (7)b. Unimaginative (6)c. Content (7)d. Approving (5)

J2:Ambitionless/Flattering (20)

a. Listless (4)b. Densely naive (4)c. Ambitionless (6)d. Flattering (6)

K:Trusting(51)

KI :Trusting/Forgiving (28)

a. Unguarded (4)b. Trusting (9)c. Innocent (5)d. Forgiving (4)e. Generous (6)

K2:Gullible/Merciful (23)

a. Totally unwary (4)b. Gullible (5)c. Guileless (6)d. Merciful (4)e. Prodigal (4)

D:Cold (37)

D,:Cold/Punitive(19)

a. Cold (5)b. Stern (8)c. Strict/punitive (6)

D2:Icy/Cruel(18)

a. Icy (5)b. Harsh (7)c. Inflexible/cruel (6)

L:Warm (37)

L,:Warm/Pardoning (19)

a. Warm (5)b. Gentle (8)c. Lenient/pardoning (6)

L2 :AH-Loving/Absolving (18)

a. All-loving (5)b. Soft (7)c. Condoning/absolving (6)

(table continued)


Table 1 (continued)

Segment, level, and subclass Opposing segment, level, and subclass

E:Hostile (44)

E,:Antagonistic/Harmful (22)

a. Antagonistic (7)b. Quarrelsome (5)c. Impolite (5)d. Harmful (5)

E2:Rancorous/Sadistic (22)

a. Rancorous (5)b. Belligerent (6)c. Rude (3)d. Sadistic (8)

M:Friendly (44)

M,:Cooperative/Helpful (22)

a. Cooperative (7)b. Agreeable (5)c. Courteous (5)d. Helpful (5)

M2:Devoted/Indulgent (22)

a. Devoted (5)b. Concurring (6)c. Hypercivil (3)d. Indulgent/consoling (8)

FrDetached (39)

F, :Aloof (20)

a. Distinterested (6)b. Distant (7)c. Preoccupied (7)

F2 :Escapistic (19)

a. Disengaged (4)b. Hermetic (7)c. Autistic/eccentric (8)

G:Inhibited (41)

G,:Taciturn(24)

a. Silent/private (7)b. Undemonstrative (4)c. Stiff/controlled (9)d. Opinionated (4)

G2:Unresponsive (17)

a. Mute (5)b. Apathetic (4)c. Catatonic (4)d. Obstinate (4)

N:Sociable (39)

N,:Outgoing(20)

a. Involved (6)b. Sociable (7)c. Extraverted (7)

N2 :Frenetically Gregarious (19)

a. Intrusive (4)b. Monophobic (7)c. Hyperactive/hyperconventional (8)

O:Exhibitionistic (41)

Oi :Spontaneous/Demonstrative (24)

a. Talkative/disclosing (7)b. Demonstrative (4)c. Casual/spontaneous (9)d. Suggestible (4)

O2:Histrionic (17)

a. Loquacious/divulging (5)b. Histrionic (4)c. Impulsive (4)d. Hypersuggestible (4)

H:Unassured (53)

H,:Self-doubting/Dependent (28)

a. Self-doubting (9)b. Dependent (5)c. Unassured (5)d. Awkward (5)e. Glum (4)

H2:Abasive/Helpless (25)

a. Abasive (9)b. Helpless (5)c. Immobilized (4)d. Bumbling (3)e. Depressed (4)

P:Assured (53)

P, :Confident/Self-Reliant (28)

a. Confident (9)b. Self-reliant (5)c. Assured (5)d. Self-composed (5)e. Cheerful (4)

P2:Arrogant/Rigidly Autonomous (25)

a. Egotistical (9)b. Rigidly autonomous (5)c. Cocky (4)d. Imperturable (3)e. Euphoric (4)

Note. Numbers of items for each segment are in parentheses.


Hence, the entire continuum of Segment Ais designated Dominant, the mild-moderatelevel is called Controlling, and the extremelevel is named Dictatorial.

Table 1 shows that the number of 1982items denning each of the 16 segments rangesfrom 37 to 53, with a median of 42.5, whereasthe number of items denning each of the 32segment levels ranges from 17 to 24, with amedian of 21.5. Notice further that each seg-ment level also is denned as the bipolar op-posite of its corresponding level on the op-posite segment (e.g., A, Controlling vs.I,:Docile; A2:Dictatorial vs. I2:Subservient).

From Table 1 it can also be seen that eachof the two levels for a particular segment isadditionally defined by three to five sub-classes. For example, Segment A:Dominantis defined by four subclasses at each of thetwo levels: A] :Controlling by a:leading/influ-encing, b:active/self-assertive, c:strong/man-aging, and d:taking charge; and A2 dictatorialby axommanding, b:overbearing, c:bossy,and d:tyrannical. For any one segment thenumber of subclasses at the two levels is iden-tical, with the equivalent letters definingmild-moderate versus extreme levels of eachsubclass (e.g., A^leading/influencing andA2a:commanding; A,b:active/self-assertiveand A2b:overbearing). Table 1 reveals that thenumber of circle items for each of the 64segment-pair subclasses ranges from 3 to 9,with a median of 5. By far the majority ofsubclasses are operationalized by from fourto seven circle items.

Not only is each level of each segment sub-class operationalized by bipolar items, buteach of the 128 subclasses is provided a cat-egory definition. For example, the four sub-classes of level AI :Controlling are defined asfollows:

a. Leading/influencing: to tell or show theway by instruction, helpful information, oradvice; to guide another in direction, course,action, opinion, and so forth; to exercisepower over the minds or behavior of others;to affect, sway, move, or impel another tosome action by direct or indirect means suchas tact, address, artifice, and so forth.

b. Active/self-assertive: to be busy or fre-quently engaged in action; to be quick anddiligent in doing, acting, performing, or

working; to put oneself forward boldly andinsistently and to resist influence.

c. Strong/managing: to engage others asmentally and morally powerful, firm, andcourageous; to direct others toward a goal;to make decisions for others, for groups, orfor organizations; to manipulate resourcesand expenditures to accomplish a purpose.

d. Taking charge: to charge, command,order, instruct, and direct others to do or notdo something; to exercise direction or re-straint over others.

The corresponding four subclasses ofA2:Dictatorial are similarly defined, as arethe subclasses for all other circle segments.

It can also be seen from Table 1 that forthe eight pairs of opposing segments thenumber of subclasses is identical, each pairbeing defined by the same bipolar items. Toillustrate, Segment I:Submissive is operation-alized by four subclasses at each level, justas is its opposite, Segment A:Dominant. Fur-thermore, the four subclasses of I;:Docile areopposites, respectively, of the four subclassesof A, Controlling (e.g., I^following/comply-ing vs. A pleading/influencing; Impassive/acquiescent vs. A ̂ active/self-assertive).Similarly, the four subclasses of ^Subser-vient are opposites, respectively, of the foursubclasses of A2:Dictatorial (e.g., I2a:servilevs. A2a:commanding; I2b:spineless vs.A2b:overbearing).

In sum, the 1982 Circle contains 350 bi-polar (700 unipolar equivalents) interper-sonal items, 3 to 9 of which operationalize1 of 64 segment-pair subclasses at either themild-moderate or extreme levels of each of16 segments. Each of the two levels of eachof the 16 segments is operationalized by 3 to5 subclasses, with the number of items defin-ing each of the 32 levels ranging from 17 to28, with a median of 21.5. Finally, each ofthe 16 segments (including both levels) is de-fined by 37 to 53 items, with a median of42.6.

Table 2 compares the 1982 Circle itemstructure to that of the four two-dimensionalcircle measures. As can be seen, only the1982 Circle provides a further breakdown ofeach segment into subclasses. Second, onlythe 1982 Circle and the ICL provide itemsfor assessing two levels of each segment, with


the 1982 Circle averaging four times as manyitems for each of the 32 levels. Also, the 1982Circle provides about five times as manyitems per segment as does any of the othermeasures.

Figure 2 depicts locations on the 1982Circle of corresponding ICL, IAS, IBI, andIMI segments based on my content analysisof the latter segment items. Because the IMIwas derived directly from the IBI, IMI seg-ments exactly overlap IBI placements.

From Figure 2 it is apparent first that allfour inventories differ from the 1982 Circlein either segment representation, placement,or both. Second, all four inventories are quiteconsistent with the 1982 Circle in represen-tation and placement of the segments thatanchor bipolar opposites on the Control andAffiliation axes, namely segments A, E, I, andM. Indeed, based on my content analysis, theICL, IBI, and IMI all "overkill" in assessmentof these circle positions by using two or moresegment scales. Also, Figure 2 provides ampleevidence that among two-dimensional circlemeasures, Dominant and Submissive seg-ments anchor the poles of the Control axis,whereas Friendly and Hostile segments definethe poles of the Affiliation dimension. Thefollowing sections summarize how each ofthe four measures contrasts with the 1982Circle.

The Interpersonal Check List (ICL)

Figure 2 confirms some of the deficienciesof the ICL noted by other writers. A majorshortcoming of the ICL is that it offers veryfew, if any, items to measure the 1982 seg-ments O and N and their opposites, F andG. In other words the ICL contains seriousmeasurement gaps in the top-right and bot-tom-left quadrants. Lorr and McNair (1965)noted these same gaps and constructed theirDetachment, Deference, Affiliation, Socia-bility, and Exhibition scales to remedy thesituation. Similarly, Wiggins (1979) devel-oped his F:Aloof, Controverted, O:Extra-verted, and N:Gregarious segments to rem-edy the problem. It follows from these seg-ment omissions that the ICL cannot measurea good one fourth of the range of two-di-mensional interpersonal actions. These gaps

Table 2Comparison of the Category and Item Structureof the 1982 Interpersonal Circle With That ofthe ICL, IBI, IMI and IAS

1982Category Circle ICL IBI IMI IAS

Number of categories measured

SegmentsLevelsSubclasses

163264(128)°

16320

15150

15150

16160

Number of items for each category

SegmentsLevelsSubclasses

Total

37-5317-283-9

350 (700)"

840

128

7-1100

140

600

90

800

128

Note. ICL = Interpersonal Check List; IBI = Interper-sonal Behavior Inventory; IMI = Impact Message In-ventory; IAS = Interpersonal Adjective Scales.a Because the 1982 items are bipolar, one pole of the 348items defines 64 subclasses for 8 segments; the other poledefines 64 corresponding subclasses for the 8 oppositesegments—a total of 128 subclasses for the 16 circle seg-ments. Similarly, each of the 350 bipolar items simul-taneously assesses two segments—hence the monopolaritem equivalent is 700 (2 X 350).

also indicate that many of the rules of com-plementarity that Leary (1957) specified forthe ICL octants are incomplete, inconsistent,and/or invalid. Furthermore, some equiva-lently labeled ICL segments represent redun-dant measurement of other 1982 segments.That is, ICL N and O actually measure the1982 L and M segments, ICL F is a redun-dant measure of 1982 E, and ICL G sharesin anchoring 1982 C.

A second major shortcoming of the ICLdescribed by Wiggins (1982) is its "lack ofpolarity between variables opposite to eachother on the circle" (p. 197). Wiggins wascareful to construct his IAS to avoid someof the spurious semantic contrasts present inseveral ICL opposites. For example, ICL Suc-cess versus Masochism was revised to produce the more genuine IAS semantic contrastof Ambitious versus Lazy; and ICL Narcis-sism versus Conformity was redefined for theIAS as Arrogant versus Unassuming. The1982 Circle, in addition to using the IAS asits major marker measure, was also carefully


lt: Docile

SUBMISSIVE

>: Subservient

Figure 2. Location of IAS, ICL, IBI, and IMI segments on the 1982 Interpersonal Circle. (IAS = Inter-personal Adjective Scales; ICL = Interpersonal Check List; IBI = Interpersonal Behavior Inventory; IMI =Impact Message Inventory.)

constructed in all aspects to avoid ungenuinesegment opposites. Indeed, as described ear-lier, the principle of opposites permeates ev-ery aspect (segment, level, subclass, and item)of the 1982 Circle.

Another problem is that when comparedto the 1982 Circle, ICL Segments P and B

are misplaced and their locations need to beexchanged. Curiously, however, the corre-sponding opposite ICL segments, H and J,are correctly placed. This is further validationof Wiggins's (1982) criticism that many ofthe ICL segment opposites are semanticallyinvalid.


On the other hand, a major strength of theICL is that it provides for systematic assess-ment of the two levels of intensity on eachof the 16 segment continua. The 1982 Circleretains this crucial circle feature and mark-edly increases the number of items used tomeasure the two levels.

The Interpersonal Adjective Scales (IAS)

From Figure 2 it is apparent that the 1982Circle and Wiggins's IAS show perfect over-lap with the exception of placement of foursegments. This is as expected because the IASwas used as the major marker measure forconstruction of the 1982 Circle. However, theIAS B and P segments—and as a result theircorresponding opposites, J and H—are ap-parently misplaced. The IAS places Ambi-tious at P and Arrogant at B, whereas the1982 Circle reverses this placement. This re-versal, in turn, necessitates reversal of theiropposites, so that IAS J:Unassuming is movedto 1982 H and IAS H:Lazy is relocated at1982 J.

The 1982 placements seem more logicallyvalid because they fit more precisely themathematical weightings of Control and Af-filiation for the respective segments. That is,Confident more precisely meets the P seg-ment weightings of +1 Friendly, +3 Domi-nant than does Ambitious—the latter moreprecisely meeting the B segment weightingsof — 1 Friendly, +3 Dominant. Also, factoranalyses of both the IBI and IMI place theCompetitive segment (the equivalent of Wig-gins's P:Ambitious) in the upper-left quad-rant consistent with P's 1982 Circle position.Furthermore, the fact that Wiggins's IASsometimes mixes assessment of the two levelsof each segment could also explain its mis-placements. That is, the eight adjectives thatoperationalize Wiggins's P:Ambitious seg-ment are all classified at the mild-moderatelevel, whereas all eight adjectives that anchorhis B:Arrogant segment describe extreme-level behaviors. Intercorrelations could re-flect this item level imbalance by pulling theposition of Wiggins's extreme-level Arrogantscale from the friendly to the hostile side ofA:Dominant.

These exchanges were made with sometrepidation because Wiggins's (1982) evi-dence for IAS circular ordering is impressive.Future research with the 1982 Circle mayindicate that my relocations are invalid andthat the four segments should be shifted backto Wiggins's positions.

The major shortcoming of the IAS is thatit does not provide systematic assessment ofboth levels (mild-moderate and extreme) ofinterpersonal acts on each of the 16 segments.My content analysis of IAS items revealedthat only 22 of the 128 IAS adjectives couldbe classified at the extreme level (cf. Kiesler,Note 3). Furthermore, these 22 items areunsystematically placed in only 7 of the 16segments. Hence, the IAS is primarily a mea-sure of the mild-moderate level with the no-table exception of the B:Arrogant segment.In contrast, the 1982 Circle provides system-atic assessment of both levels for each of the16 segments and also expands the content ofeach segment at both levels to subsumehighly correlated but distinct subclasses.

Kiesler (Note 3) also shows the IASA:Dominant segment is not homogeneousbecause five of the eight items anchor the1982 A segment, whereas the other three aremore correctly classified under 1982 P. Sim-ilarly, two of the eight items from both IASI:Submissive and G:Introverted are more cor-rectly classified under 1982 H. Finally, IASN:Gregarious seems to underrepresent 1982N because one half of the IAS N items aremore correctly classified under 1982 M.

The Interpersonal Behavior Inventory (IBI)and the Impact Message Inventory (IMI)

Because the IMI was derived from the IBI,the following observations regarding the IBIequally characterize the IMI. From Figure 2it is apparent first that the IBI contains twomajor assessment gaps: at D in the top-leftquadrant and at K in the bottom-right quad-rant. The IBI and IMI thus do not measurethe 1982 D:Cold and K:Trusting segments ofinterpersonal actions.

Also, content analysis (Kiesler, Note 3) re-veals that four other 1982 segments are only


minimally or poorly measured by the IBI—1982 E:Hostile subsumes only five IBI Hos-tility items; G:Inhibited contains only five IBIInhibition and Detachment items; and theIBI Exhibition items are divided between1982 Segments O:Exhibitionistic (five items)and P:Assured (three items). Hence, in ad-dition to being completely unable to measure1982 Segments D and K, the IBI and IMIprovide only minimal or partial assessmentof Segments E, G, O, and P.

Figure 2 also shows that the IBI and IMI"overkill" in assessment of four other 1982segments. This is especially the case for 1982G (which includes two IBI segments: De-tachment and Inhibition), for 1982 H (whichsubsumes two IBI segments: Abasiveness andSuccorance), and for 1982 M (which containstwo IBI segments: Agreeableness and Nur-turance). It is also the case to a lesser degreefor 1982 B, which, besides IBI Competitive-ness, includes four IBI Hostility items.

Another important shortcoming of the IBIand IMI is that their segments unsystemati-cally mix items that measure mild-moderateand extreme levels of interpersonal acts. Al-most one half (61) of the 140 IBI items areclassified at the extreme level. Furthermore,mixtures for the various segments are incon-sistent, and neither the IBI nor the IMI pro-vides separate scores for the two levels.

An important future task is to revise andexpand the IMI to a form where it can sys-tematically assess both levels of the 1982Circle segments. As Wiggins (1982) observes,a comprehensive interpersonal assessmentbattery needs to include a measure that tar-gets the covert or inner engagements expe-rienced by interpersonal interactants. TheIMI uniquely performs this assessment task,albiet inadequately.

Summary

In sum, the ICL, IAS, IBI, and IMI allexhibit both major and minor circle defi-ciencies. These inadequacies, in turn, (a) pre-clude valid assessment of the full range ofinterpersonal actions in their normal and ab-normal spheres, (b) prohibit derivation andtest of the laws of interpersonal complemen-tarity, (c) delay answers to exact placementon the circle of DSM-III (American Psychi-

atric Association, 1980) personality and otherdisorders, and (d) delay construction of dif-ferential interpersonal interventions for ho-mogeneous groups of psychotherapy clientswhose rigid interpersonal styles are definedat various segments or octants of the inter-personal circle.

It is my contention that the 1982 Circle,in contrast, provides the comprehensive andprecise definition and operationalization ofthe 16 circle segments, which will permitheuristic pursuit of the important problemsof personality and psychotherapy. The secondsection of this article attempts to demon-strate this utility by deriving systematicallythe laws of complementarity inherent in the1982 Circle.

Before leaving this section, however, a briefdiscussion of Benjamin's (1974, Note 4)Structural Analysis of Social Behavior (SASB)measures is necessary. I did not include theSASB before because it is not in the form ofa two-dimensional circle organized aroundthe axes of Control and Affiliation.

Instead, Benjamin's SASB model is builtupon two circumplexes designed to integratethe circles of Leary (1957) and Schaefer(1959). One circumplex, the parentlike oractive plane, is constructed around the ver-tical axis of Power versus Autonomy. The sec-ond, the childlike or reaction plane, is con-structed around the vertical axis of Individ-ualism versus Submission. This separation ofthe two planes reflects Benjamin's assump-tion that Autonomy is the opposite of Power,whereas Submission is the complement ofPower. Accordingly, her second plane depictsthe complements (reactions and their oppo-sites) to the interpersonal behaviors (actionsand their opposites) arranged on the circum-plex of her first plane.

Comparison of Benjamin's two diagonalplanes (reflecting three underlying dimen-sions) to the two-dimensional framework ofthe 1982 Circle provides a few conclusions.First, the lower halves of her two planes con-tain the four quadrants of the 1982 Circle.The lower half of Plane 1 assesses "hostile-power" and "friendly-influence," and thelower half of Plane 2 measures "hostile-com-ply" and "friendly-accept." Second, the re-maining upper halves of her two planes de-scribe interpersonal behaviors that seem quite


troublesome to place anywhere within the1982 Circle. These latter quadrants may mea-sure behaviors restricted more to real parent-child interactions and, hence, occurring lessfrequently among adult interactants. Third,until very recently Benjamin's two planeswere denned solely by the 36 individual itemsaround each diagonal. As Wiggins (1982)observes, Benjamin "tends to analyze all 36variables on each plane rather than reducingitems to a smaller set of octant or sixteenthscales" (p. 195). The unfortunate result isthat SASB items are not grouped into thefamiliar 16 segments with corresponding cat-egory definitions, although in her 1980 ver-sion Benjamin (Note 4) provides octant clus-ters for both planes. Fourth, the SASB doesnot provide systematic categorization of themild-moderate versus extreme levels of in-terpersonal acts on the two planes. Fifth, asWiggins (1982) argues, it is difficult at presentto determine with any clarity whether Ben-jamin's two planes do in fact form separatecircumplexes.

Nevertheless, Benjamin's model has dem-onstrated its fine-tuned power in multipleapplications to the psychotherapy situation(Benjamin, 1977, 1979a, 1979b, 1982). HerSASB system has been "by far, to date, themost scientifically rigorous and clinically as-tute model published" (McLemore & Hart,1982, p. 233) and "the most detailed, clini-cally rich, ambitious, and conceptually de-manding of all contemporary models" (Wig-gins, 1982, p. 193).

A central issue underlying this SASB dis-cussion is whether two or three dimensionsmore validly represent the basic structure ofinterpersonal behavior. Two-factor structureis supported by findings for the ICL and espe-cially for the IAS. Inconsistent support comesfrom findings for the IBI and IMI. And, asWiggins (1982) concludes, despite the largeamounts of evidence presented by Benjamin,it is difficult to answer with any degree ofcertainty the questions of whether her twoSASB planes "do in fact form separate cir-cumplexes, and whether the complementaryrelations between these two planes are asspecified" (p. 195). Before this crucial theo-retical question can be answered more defin-itively, the most essential requirement seemsto be availability of two-dimensional circle

measures that meet all the logical and em-pirical criteria outlined earlier and incorpo-rated into the 1982 Circle taxonomy.

Interpersonal Complementarity Revisited

Since Sullivan (1953), a central constructof interpersonal theory has been the reci-procity or complementarity governing theexchanges of human interactants. Surpris-ingly, little systematic development of thisconstruct has occurred in subsequent inter-personal theory, however. If complementarityis a valid construct, important and multipleunderstandings and predictions follow for thewide range of human social behavior includ-ing interactions between client and therapist.The purpose of this second section is to pro-vide a systematic statement of propositionsof complementarity inherent in the interper-sonal circle in general and in the 1982 Circlein particular.

Before proceeding with this task, however,two other traditions that have addressed thenotion of interpersonal reciprocity need tobe mentioned. The first has been provided byinteractional communication theorists (Bate-son, 1958; Berne, 1964; Goffman, 1967;Haley, 1963; Jackson, 1959; Lederer & Jack-son, 1968; Ruesch & Bateson, 1951; Sluzki& Beavin, 1977; Watzlawick, Beavin, & Jack-son, 1967; Watzlawick & Weakland, 1977;Winch, 1958). These formulations, in con-trast to those deriving from the interpersonalcircle, have concentrated on reciprocity as itoperates on the Control dimension only. Thatis, reciprocity in an interpersonal exchangerepresents the constant struggle by each per-son to control what sort of power relationshipis to exist between them (Haley, 1963).

Bateson (1958) was the first to use theterms complementary and symmetrical todescribe dissimilar and similar control be-haviors among dyadic interactants. Jacksonand Haley (Jackson, 1959) elaborated Bate-son's notions in the context of their controltheory as follows: "In a complementary re-lationship the two people are of equal statusin the sense that one appears to be in thesuperior position, meaning that he initiatesaction and the other appears to follow thataction." In contrast, a "symmetrical relation-ship is one between two people who behave


as if they have equal status. Each person ex-hibits the right to initiate action, criticize theother, offer advice, and so on" (pp. 126-127).Furthermore, in the complementary rela-tionship the person culturally or contextuallyrecognized as being superior is in the "one-up" or primary position, whereas the personrecognized as being inferior is in the "one-down" or secondary position (Haley, 1963).Lederer and Jackson (1968) added a thirdtype of reciprocity, a parallel relationship,wherein control not only tends to have anequal distribution among interactants butalso flows easily and alternatingly from A toB to A as the situation changes and dictates.

A second alternative tradition regardingcomplementarity comes from Schutz's (1958)three-dimensional theory of interpersonalbehavior, and from research with his Fun-damental Interpersonal Relations Orienta-tion Scale (FIRO-B). For Schutz, three typesof interpersonal reciprocity or compatibilityare possible as the result of the operation ofthree classes of needs: inclusion, control, andaffection, (a) Reciprocal compatibility existswhen each participant's level of "expressed"behavior matches the other's level of "wanted"behavior for each of the three need areas, (b)Originator compatibility describes reciproc-ity in regard to who originates and who re-ceives behaviors for each of the three needareas. If both participants prefer to originatebehaviors, "competitive" incompatibility ex-ists; if both prefer to receive, "apathetic" in-compatibility occurs, (c) Interchange com-patibility designates the degree to which bothparticipants rank similarly the respective im-portance of the three need areas. The FIRO-B measures each of these three types of com-patibility.

Now we can bring our focus to bear on thethird tradition, interpersonal theory, whichoffers notions of reciprocity directly derivablefrom the interpersonal circle itself.

Sullivan (1953) first offered the notion ofinterpersonal reciprocity with his "theoremof reciprocal emotion," which states that"integration in an interpersonal situation isa process in which (1) complementary needsare resolved (or aggravated); (2) reciprocalpatterns of activity are developed (or disin-tegrated); and (3) foresight of satisfaction (orrebuff) of similar needs is facilitated" (p.

129). Furthermore, the outcome of a partic-ular integrative attempt may be either con-junctive (leading to inclusion or further in-teraction) or disjunctive (leading to exclusionor avoidance of further interaction). Theseintegrative attempts occur in the context ofa dyadic or larger social system and, in thecase of the dyad, neither Person A's needs norPerson B's needs alone determine the out-come. Interactants' interpersonal needs al-ways seek conjoint expression and resolution;hence, interpersonal behavior can be under-stood only from a systems perspective. Thisfocus on two-person mutual influence or bi-directional causality represents a fundamen-tal novelty of Sullivan's formulations.

The broadest notion of reciprocity or com-plementarity, thus, is that our interpersonalactions are designed to invite, pull, elicit,draw, entice, or evoke "restricted classes" ofreactions from persons with whom we inter-act, especially from significant others. Reac-tions by others to our acts are not random,nor are they likely to include the entire rangeof possible reactions. Instead, reactions to ustend to be restricted to a relatively narrowrange of interpersonal responses. At mostlyunaware and automatic levels, our actions aredesigned to push or force others to respondin ways that are complementary to our actsand that confirm our self-definitions.

For Danziger (1976), "redundancy" is thekey aspect of circular causality inherent inthis notion of reciprocity or complementar-ity. That is, human transactions over timeshow consistencies and regularities; dyadsspin out an "orderliness in the sequence ofinteraction" (p. 184). The sequence containsa series of reciprocal exchanges that showsa significant deviation from chance ordering,which manifests a patterned order.

Unfortunately, Sullivan (1953) offered verylittle systematic amplification of this theo-rem. The surviving general notion of com-plementarity was that actions of human par-ticipants are redundantly interrelated (i.e.,have patterned regularity) in some mannerover the sequence of transactions.

Leary (1957) provided a more systematicrestatement of Sullivan's theory. Equally im-portant, he offered an operational definitionof complementarity within his interpersonalcircle and by the ICL developed by LaForge


and Suczek (1955). Leary defined his "prin-ciple of reciprocal interpersonal relations" asfollows: "Interpersonal reflexes tend (with aprobability significantly greater than chance)to initiate or invite reciprocal interpersonalresponses from the 'other' person in the in-teraction that lead to a repetition of the orig-inal reflex" (p. 123). Furthermore, a personprovokes responses from others that reinforcethe person's original actions, a feature Learycalled "the reinforcing quality of social in-teraction" (p. 123).

Throughout his book Leary highlightedthis "reciprocal nature of social interaction,the reflex way in which humans tailor theirresponses to others, and the automatic wayin which they force others to react to them"(p. 83); this "reciprocal process by whichhuman beings tend to pull from others re-sponses that tend to maintain their limitedsecurity operations" (p. 112); and this "au-tomatic process by which we pull certain re-actions from others and, in turn, respondwith a limited set of behaviors" (p. 156).

Surprisingly, Leary did not explicitly de-fine the rules of complementarity, althoughthey were clearly implicit in his interpersonalcircle and were explicitly applied to person-ality and clinical groups in the second halfof his volume. Leary did clearly identify twomajor dimensions or rnotivations that formedthe structure of interpersonal behavior (andthe interpersonal circle)—namely, domi-nance-submission and love-hate. Impliedthroughout his volume was the fact that thevertical axis of his circle reflected the rulethat dominant reflexes elicit submissive re-actions, and vice versa, whereas on the hor-izontal axis love pulls love, and hate elicitshate. These basic rules were illustrated inLeary's detailed descriptions of eight inter-personal styles, or personalities, and theirpsychiatric equivalents, which could be clas-sified at the various octants of his circle. Inthese descriptions, interpersonal styles on theleft side of the circle pulled complementaryoctants residing in the other quadrant of theleft side (e.g., FG pulled CD), whereas right-side octants pulled octants falling in theother quadrant on the right side (e.g., NOpulled KL).

It was left to Carson (1969) to articulatethe principle of complementarity implicitly

present and explicitly applied throughoutLeary's volume. Carson first restated Leary'sprinciple of reciprocal interpersonal relationsas follows: "The purpose of interpersonal be-havior, in terms of its security-maintenancefunctions, is to induce from the other personbehavior that is complementary to the be-havior proffered" (p. 112). Adopting Leary'stwo basic circle dimensions, Carson observedthat interpersonal behaviors convey implicitmessages that give or deny love or status tothe self or to the other person. Hence, "aninterpersonal act represents, in part, a promptor 'bid' to elicit response behavior fallingwithin a certain range of the interpersonalcircle" (p. 115), and "the adoption of a par-ticular interpersonal stance in large partserves the function of producing a particularstance in the other person" (p. 143).

Carson also offered an important explicitdefinition of the complementarity that gov-erns interpersonal relations. "Generallyspeaking, complementarity occurs on the ba-sis of reciprocity in respect to the dominance-submission axis (dominance tends to inducesubmission, and vice versa), and on the basisof correspondence in respect to the hate-loveaxis (hate induces hate, and love induceslove)" (p. 112). Thus, Carson designatedcomplementarity as the more general termand defined the two subsidiary concepts ofreciprocity and correspondence. Also, echo-ing Leary, he reiterated the assumption thata "complementary interaction is in itselfmutually rewarding to at least some degree,probably by way of enhancing the securityof both participants" (p. 145).

In addition, Carson was the first to specifythe contrasting but crucial anticomplemen-tary and noncomplementary relationships asthey operate on the quadrants of Leary's in-terpersonal circle. Carson renamed Leary'shorizontal axis "hostile-friendly"; the result-ing quadrants became "hostile-dominant"(HD), "friendly-dominant" (FD), "hostile-submissive" (HS), and "friendly-submissive"(FS). "When a person 'offers' behavior fallingwithin any of the quadrants of the interper-sonal circle, he is, in effect, 'inviting' the otherperson to adopt a complementary stance inrespect to both of the principal dimensionsof the circle" (p. 147). However, the partici-pant's reactions can take one of three distinct


forms, (a) "If the other person 'accepts' bothaspects of the invitation or proposal" (p.147), he adopts the complementary position(i.e., HD pulls HS, HS pulls HD, FD pullsFS, FS pulls FD). (b) If the other person ac-cepts only one of the component messages,he adopts a noncomplementary position (i.e.,HD pulls FS, FS pulls HD, FD pulls HS, HSpulls FD). (c) Finally, if the other person re-jects both aspects of the invitation or pro-posal, he adopts an anticomplementary po-sition (i.e., HD pulls FD, FD pulls HD, HSpulls FS, FS pulls HS). In the remainder ofthis article I will refer to Carson's noncom-plementary response as the acomplementaryresponse, and use "noncomplementary" tosubsume both acomplementary and anti-complementary responses.

According to Carson, a complementaryinteraction is in itself mutually rewarding toat least some degree. Similarly, the acomple-mentary interaction, "although lacking per-fect complementarity, nevertheless leavesopen a channel of potential transaction andnegotiation" (p. 147) and, hence, is poten-tially still mutually rewarding. In contrast,the anticomplementary interaction, becauseit involves rejection of the person's bid forboth love and status, embodies "a relativelycomplete repudiation" (p. 148) of the per-son's proffered self-definition. As a result,"it leaves very little ground upon which toeffect a sustained engagement" (pp. 147-148)and, potentially, is minimally rewarding oraversive.

It is necessary to point out that neitherLeary nor Carson specified another type ofacomplementarity inherent in the interper-sonal circle, namely, what occurs when Per-son A and Person B fall at identical circlequadrants or segments. That is, a hostile-dominant person can interact with anotherhostile-dominant person, a friendly-submis-sive person with another friendly-submissiveperson, and so on. These interactions involv-ing identical styles are acomplementary be-cause complementarity exists on one axis butnot on the other (e.g., in the case of FD-FD,correspondence exists on "hostile-friendly,"but nonreciprocity exists on "dominance-submission"). Hence, from this point on Iwill distinguish two types of acomplemen-tarity: isomorphic, or those transactions in

which interactants show identical styles at thesegment, octant, or quadrant levels, andsemimorphic, or those other acomplemen-tary transactions that show complementarityon one but not both circle axes. As we willsee, the semimorphic acomplementary re-sponse always falls at segments directly op-posite on the circle.

It is interesting to note that these comple-mentary and noncomplementary character-izations represent an important departure ofinterpersonal theory from the commonsensenotions that either "opposites" or "likes" at-tract. In interpersonal theory, likes (iso-morphics) and opposites (semimorphics) onlypartially attract, whereas they also partiallyrepel. Rather, complementaries attract totallyand anticomplementaries repel totally.

Benjamin (1974), applying Carson's com-plementarity and noncomplementarity con-cepts to development of her three-dimen-sional SASB model, offered the first circledefinition of the concept of opposites. Forher, opposites are defined as segments thatreflect the same proportions of the underlyingdimensions (Control and Affiliation for two-dimensional measures), except that the signsare changed. For example, Segment B is theopposite of Segment J because B is definedmathematically as -1 Affiliation, +3 Con-trol, whereas J is defined as +1 Affiliation,—3 Control. The principle of opposites (cf.also Wiggins, 1982) represents a circle prop-erty essential to precise definition of comple-mentarity, acomplementarity, or anticomple-mentarity. In order for the mathematics ofEuclidian space to apply, each of the 16 seg-ments needs to define acts behaviorally andsemantically opposite to the segment at theother end of a particular circle diameter.Valid laws of complementarity cannot bederived from the interpersonal circle unlessit contains, as does the 1982 Circle, eightpairs of exact segment opposites that defineclearly contrasting classes of interpersonalacts.

Having completed this review of interper-sonal circle notions of complementarity, Iwill now summarize these formulations withthree theoretical propositions.

PROPOSITION 1. A person's interpersonalactions tend (with a probability significantly


greater than chance) to initiate, invite, orevoke from an interactant complementary re-sponses that lead to a repetition of the per-son 's original actions.

Interpersonal behaviors, in a relatively un-aware, automatic, and unintended fashiontend to invite, elicit, pull, draw, or enticefrom interactants restricted classes of reac-tions that are reinforcing of, and consistentwith, a person's proffered self-definition. Ifcomplementary reactions are not forthcom-ing from interactants, the relationship willeither not endure or it will be altered in sucha manner that complementarity is estab-lished.

This proposition is a restatement of Leary's(1957) principle of reciprocal interpersonalrelations. It also describes the "command"level (in contrast to the "report" level) ofhuman communication described by Bateson(1958) and Watzlawick, Beavin, and Jackson(1967).

From his "communications analytic" po-sition, Beier (1966) labeled a particular in-terpersonal action the "evoking message."According to Beier, the evoking message im-poses a condition of emotional engagementon a decoder, and, as a result, the decodercountercommunicates as the encoder wishes,without being aware of complying. The en-coder also is unaware that he or she has im-posed a condition or sent a command mes-sage and, consequently, obtains responsesfrom decoders for which he or she cannotaccount, even though the responses were self-elicited. Beier contrasts the evoking messageto the "persuasive message" in which the en-coder is aware of what he or she wants fromor is imposing on the decoder. Beier (1966)notes:

The evoking message is probably one of the basic toolsused by individuals to maintain their consistency of per-sonality. With this message an individual can elicit re-sponses without being aware that he is responsible fordoing so. To a certain extent he can create his environ-ment without feeling that he is accountable for the re-sponses which come his way. (p. 13)

Kiesler (1979, 1982a; Kiesler, Bernstein,& Anchin, 1976; Kiesler, Note 5) placesBeier's evoking message and its decoderequivalent, the impact message," in centralpositions in his interpersonal communication

theory for psychopathology and psychother-apy. For Kiesler, the impact message refersto the distinctive pattern, the particular com-plex of covert, internal engagements (feelings,cognitions, fantasies) an interactant recur-rently experiences as the direct effects of aperson's interpersonal behavior. Kiesler andhis colleagues developed the IMI to opera-tionalize the domain of impact messages inpersonality and psychotherapy.

PROPOSITION 2. For interpersonal behavioras operationalized by the two-dimensionalinterpersonal circle, complementarity occurson the basis of (a) reciprocity in respect to theControl dimension or axis (dominance pullssubmission, submission pulls dominance),and (b) correspondence in respect to the Af-filiation dimension (hostility pulls hostility,friendliness pulls friendliness).

This proposition is a restatement of Car-son's (1969). A growing body of empiricalresearch confirms these two rules of comple-mentarity (Beery, 1970; Berzins, 1977, Note6;Celani, 1974; Crowder, 1972; Cutler, 1958;DeVogue & Beck, 1978; Dietzel & Abeles,1975; Gaminsky & Fairwell, 1966; Heller,Myers, & Kline, 1963; Mueller, 1969; Mueller& Dilling, 1968, 1969; Raush, Dittman, &Taylor, 1959; Raush, Farbman, & Llewellyn,1960; Rice, 1970; Smelser, 1961; Swensen,1967; Luborsky & Singer, Note 7) with somemixed results (Edquist, 1973; Malone, 1975;Shannon & Guerney, 1973).

This proposition has significant implica-tions for redefinition of "social reinforce-ment" in behavior theory and therapy, asDe Vogue and Beck (1978) and Kiesler, Bern-stein and Anchin (1976) have argued. Forexample, De Vogue and Beck observe:

Only in friendly dominant/friendly submissive dyads. . . would social "reinforcement" in the form of praiseand approval have its maximum effect. In any nonrecip-rocal dyad (i.e., in which the subject attempted to utilizehostile dominance, hostile submission, or friendly dom-inance) we could predict less frequent use of the targetresponse by the subject than in this reciprocal dyad(where the subject displayed a friendly submissive tactic). . . It could also be predicted that within any of thesenonreciprocal dyads, but especially when the subjectattempts to use hostile dominance vis-a-vis the rein-forcer, the subject may show a decrease in the targetresponse in order to avoid the aversive stimulation ofpraise and approval, (p. 221)


[Further, for those people] employing tactics for whichthe reciprocal is unfriendliness, coldness, rejection, de-pendency, or deference from others, praise and approvalare aversive stimuli. In fact, for the latter group a rep-licated finding is that the removal of praise and approvalhas functioned as a negative reinforcer. In short . . .those most likely to experience praise and approval asreinforcing are those biased toward friendly submission,while those least likely are those limited to hostile dom-inance . . . This transactional pattern suggests the im-portance of utilizing an interactional theory to enhanceformulations derived from conditioning-based models,(p. 228)

PROPOSITION 3. For interpersonal behavioras operationalized by the two-dimensionalinterpersonal circle: (a) Complementarity ex-ists among interactants when Respondent Breacts to Person A with interpersonal acts re-ciprocal in terms of Control and correspond-ing in terms of Affiliation; (b) anticomple-mentarity exists when Respondent B reactsto Person A with behavior both nonreciprocalin terms of Control and noncorresponding interms of Affiliation; (c) acomplementarity ex-ists among interactants when Respondent Breacts to Person A with actions either recip-rocal on Control or corresponding on Affili-ation, but not both; (d) isomorphic acomple-mentarity exists when Respondent B reactsfrom circle segments identical to those usedby Person A; and (e) semimorphic acomple-mentarity exists when Respondent B reactsfrom circle segments directly opposite to thoseused by Person A.

We can now illustrate, using the 1982 Cir-cle, the three propositions we have developedso far. Figure 3 depicts interpersonal com-plementarity as it operates among the quad-rants and among the segments of the 1982Circle. It is apparent from Figure 3 that thecomplementary response always occurs ver-tically within the circle, and always withinthe right or within the left halves of the circle.This is necessarily the case as the result ofthe rule of correspondence (hostility pullshostility on the left side, friendliness pullsfriendliness on the right).2

Figure 4 shows anticomplementarity as itworks among the quadrants and among thesegments of the 1982 Circle. It can be seenthat the anticomplementary response alwaysoccurs horizontally within the circle and al-ways within the top or within the bottomhalves of the circle. This is necessarily the

Figure 3. Complementary quadrants and segments ofthe 1982 Interpersonal Circle.

case because the anticomplementary re-sponse always entails noncorrespondence(hostility is responded to with friendliness,friendliness with hostility) and nonreciproc-ity (dominance is responded to with domi-nance, submission with submission). As Ben-jamin (Note 8) observes, "the concepts ofopposition and complementarity can becombined to yield a prediction for antithesis(anticomplementarity): it is found by movingto the opposite and then to the 'complement'of the opposite" (p. 8).

Figure 5 shows possibilities of semi-morphic and isomorphic acomplementarityas they operate among the quadrants andamong the segments of the 1982 Circle. Ascan be seen, all instances of semimorphic

2 Oden (1976) recently constructed a parlor game,TAG (Transactional Awareness Game), that providessimilarly precise complementary circle predictions butis based on Leary's (1957) segment definitions.


Figure 4, Anticomplementary quadrants and segmentsof the 1982 Interpersonal Circle.

acomplementarity occur directly across thediameters at the opposite segments, whereasall instances of isomorphic acomplementar-ity fall at identical circle segments.

It is important to note that two other de-pictions of each of Figures 3, 4, and 5 arenecessary for precise specification of com-plementarity and noncomplementarity onthe 1982 Circle. These illustrations would beidentical to the respective figures except thatone would show the 16 segments using theircorresponding mild-moderate i labels and theother would show the 16 segments using theircorresponding extreme2 labels (cf. Figure 1).These additional figures reflect another prop-osition of complementarity that needs to beadded at this point.

PROPOSITION 4. Interpersonal complemen-tarity and noncomplementarity operate pre-cisely only within the same level or intensityof behavior. That is, interpersonal actions at

a particular level of intensity tend (with aprobability significantly greater than chance)to initiate, invite, or evoke from interactantscomplementary responses at the equivalentlevel of intensity (mild-moderate actions pullmild-moderate complementary responses,extreme acts pull extreme complementary re-sponses).

It follows from Proposition 4 that it is aconceptual error to mix segment levels instudying interpersonal complementarity. Forexample, A] iControlling pulls Ii :Docile, notI2 :Subservient. Likewise, A2: Dictatorial pulls12:Subservient, not Ii:Docile, and so on. Thisproposition also underscores again the vitalnecessity of operationalizing at least two lev-els of acts for each of the 16 circle segments.

Before leaving these initial propositions ofcomplementarity, we need to illustrate themalso at the octant level, which will be done

Figure 5. Acomplementary quadrants and segments ofthe 1982 Interpersonal Circle (solid lines = semimorphic,broken lines = isomorphic).

204 DONALD J. K.IESLER

after a few necessary introductory observa-tions. First, the last half of Leary's (1957)volume details the workings of complemen-tarity (but not of anticomplementarity oracomplcmentarity) at the octant level for hiseight personality and abnormal groups. AsTable 3 shows, however, Leary did not applycircle complementarity either consistently orprecisely. The precise complementary re-sponse (the column labeled "1982 Circle")for each of the octants involves two segmentsonly (e.g., AP pulls IJ, BC pulls GH), whereasLeary defines the complementary responsewith as many as four different segments, anddefines complementarity validity for only 3of the 8 octants (11, 77, 88).

Second, it needs to be emphasized thatactually 16 different octants are definable forthe interpersonal circle (the total list resultsfrom combining the last and the third-from-last columns in Table 3)—not merely theeight elaborated by Leary and others. Thisis the case because interpersonal acts at anyone segment (e.g., C) can blend with acts ofthe highly correlated adjacent segment oneither side (i.e., B as well as D). Each of the16 segments represents a conceptually "pure"class of interpersonal acts that never occursalone in a given person. An individual's be-havior, at a minimum, is a blend of acts fromat least two adjacent segments and may com-bine a triad of segments (a peak segment plusone segment from each side of the peak). Apressing issue for interpersonal theory andresearch concerns exactly how large a sliceof circle segments is encompassed by an in-

dividual's behavior. Furthermore, accordingto the interpersonal "rigidity" definition ofmaladjustment (Leary, 1957), the slice is ofnecessity smaller for the maladjusted personthan it is for the more normal individual.

Table 4 lists each of the 16 possible octantsof the 1982 Circle and its complementary,anticomplementary, semimorphic-acomple-mentary, and isomorphic-acomplementaryresponses. This table can be used to findquickly the complementary and noncomple-mentary predictions for any one of the 16octants. For example, assume that a partic-ular person is defined at Octant DE:Cold-Hostile. We can then predict that this personis likely to evoke from others the comple-mentary response at Octant EF:Hostile-De-tached (cold pulls detached, hostile pulls hos-tile). Anticomplementary interactants, whomthe person tends to avoid, are characterizedat Octant MN:Friendly-Sociable. Acomple-mentary interactants who are semimorphic(opposites) will fall at Octant LM:Warm-Friendly, and those who are isomorphic(likes) will be characterized at the identicaloctant, DE:Cold-Hostile. The person willtend to experience approach-avoidance ten-dencies with both acomplementary types.Similar exact predictions can be made foreach of the other 15 octants.

Wiggins (1982) offers an alternative octantformulation of complementarity based onthe work of Foa (1961, 1965, 1966; Foa &Foa, 1974) and Carson (1969, 1979) and re-flecting the hypothesis that interpersonal be-havior is structured around three facets: di-

Table 3Comparison of Theoretically Precise 1982 Interpersonal Circle With Leary's (1957) List ofComplementary Octants for the Eight Personality Types and Their Psychiatric Equivalents

Complementary segments

Octant Personality Psychiatric equivalent Segments Leary 1982 Circle

1122334455

66

7788

AutocraticNarcissisticSadisticDistrustfulMasochistic

Dependent

OverconventionalHypernormal

NoneNonePsychopathSchizoidDepressiveObsessiveAnxiousPhobicHystericPsychosomatic

PABCDEFG

HIJK

LMNO

IJGH, IJF, GHB, CD

B, CD, EN, OP, A

MNKL

IJGHEFCD

AB

OPMNKL


Table 4Predicted Complementary, Anticomplementary, and Acomplcmentary Octants for Each of the 16Interpersonal Style Octants of the 1982 Interpersonal Circle"

Complementary and noncomplementary octants

1982 Circle Complementary AnticomplementarySemimorphic

acomplementary

PA:Assured-DominantBGCompetitive-

Mistrusting

DE:Cold-HostileFG:Detached-Inhibited

HI:Unassured-SubmissiveJK:Deferent-TrustingLM: Warm-FriendlyNO:Sociable-

Exhibitionistic

IJ:Submissive-DeferentGH:Inhibited-Unassured

EF:Hostile-DetachedCD:Mistrusting-Cold

AB:Dominant-CompetitiveOP:Exhibitionistic-AssuredMNiFriendly-SociableKL:Trusting-Warm

AB:Dominant-Competitive HI:Unassured-SubmissiveOP:Exhibitionistic-Assured JK:Deferent-Trusting

AB:Dominant-Competitive HI:Unassured-SubmissiveCD:Mistrusting-Cold FG:Detached-Inhibited

EF:Hostile-DetachedGH:Inhibited-Unassured

IJ:Submissive-DeferentKL:Trusting-Warm

MN:Friendly-SociableOP:Exhibitionistic-Assured

DE:Cold-HostileBCCompetitive-

MistrustingPA:Assured-DominantNO:Sociable-

ExhibitionisticLMiWarm-FriendlyJK:Deferent-Trusting

MN: Friendly-SociableKL:Trusting-Warm

IJ: Submissive-DeferentGH:Inhibited-UnassuredEF:Hostile-DetachedCD:Mistrusting-Cold

PA: Assured-DominantNO:Sociable-

ExhibitionisticLM:Warm-FriendlyJK: Deferent-Trusting

HI:Unassured-SubmissiveFGrDetached-Inhibited

DE:Cold-HostileBGCompetitive-

Mistrusting

LM:Warm-FriendlyNO:Sociable-

ExibitionisticPA: Assured-DominantBC:Competitive-MistrustingDE:Cold-HostileFG:Detached-Inhibitcd

IJ:Submissive-DeferentKL:Trusting-Warm

MN:Friendly-SociableOP:Exhibitionistic-Assured

AB:Dominant-CompetitiveCD:Mistrusting-Cold

EF:Hostile-DetachedGH:Inhibited-Unassured

8 Corresponding isomorphic-acomplementary octants are identical to the original 1982 Interpersonal Circle octants(e.g., PA for PA, BC for BC).

rectionality (giving vs. taking, accepting vs.rejecting), object (self, other), and resource(status, love). Wiggins (1982) aligns the eightcombinations of these three facets with therespective octants of his IAS circle, statingthat "any given interpersonal variable maybe reduced to its underlying facet elementswith reference to its profile coding on theseelements" (p. 215). He derives from the facetstructure of each octant the specific "kindsof dyads formed when an initial definitionof an interpersonal situation is accepted bythe other participant" (p. 216)—that is, thespecific complementary pairs of octantsderived from his facet analysis. For exam-ple, Wiggins's complementary response toPA:Ambitious—Dominant behavior is LM:Warm-Agreeable, a response in which LMgrants love but not status to himself whilegranting both love and status to PA, as re-quested.

Wiggins's predictions for complementarityoffer a clear theoretical alternative to the oc-tant predictions of the 1982 Circle listed inTable 4. In no case do the corresponding oc-tant predictions agree (for the example citedbefore, the 1982 Circle PA pulls IJ, not LM;etc.). Also Wiggins's formulations need to beexpanded to include all 16 octants. Neverthe-less, these contrasting definitions of octantcomplementarity provide competing guidesto future interpersonal theory and research.

Now we need to add seven other propo-sitions of complementarity that have notbeen articulated at all by previous writers.

PROPOSITION 5. A given instance of thecomplementary response consists of a two-stage sequence occurring rapidly in an inter-actant: (a) a covert response, labeled the "im-pact message," and (b) the subsequent overtaction, labeled the "complementary re-sponse."


Kiesler (1979, 1982a; Kiesler, Bernstein,& Anchin, 1976; Kiesler et al., 1976; Perkinset al., 1979; Kiesler, Note 5) describes theimpact message as including those emotions,action tendencies, cognitive attributions, andfantasies that represent Interactant B's covertengagements, or pulls, in his or her transac-tions with Person A. These classes of impactsall occur internally in B's experience andmediate in large part his or her subsequentovertly expressed complementary response.The impact message, thus, is the distinctivecomplementary pattern of covert engagementexperienced by one interactant in responseto another. In contrast, the actions (wordsand nonverbal behavior) Interactant B ac-tually emits in response to Person A comprisethe traditionally labeled complementary re-sponse.

To illustrate, Interactant B starts a trans-action with Person A, whose behaviors aredefined at Octant HI:Unassured-Submissive.As their transaction proceeds, B increasinglyexperiences the covert first stage of comple-mentarity by registering complementary im-pacts to HI, such as feeling "superior to him"and "frustrated that he won't defend his po-sition"; action tendencies, such as "I shouldbe very gentle with him" and "I could tellhim anything and he would agree"; and cog-nition attributions such as "he thinks he isinadequate" and "he would accept whateverI said." As B continues to experience thesecomplementary internal engagements, hisovert actions, the second-stage complemen-tary response, increasingly manifest behav-iors from the complementary octant,AB: Dominant-Competitive.

Future theory and research face the con-siderable task of determining to what preciseextent within live transactions the proposi-tions of complementarity apply isomorphi-cally to both of these stages (i.e., to the first-stage covert complementary response or im-pact message and to the second-stage overtcomplementary response).

PROPOSITION 6. The more extreme andrigid (maladjusted) the interpersonal style ofInteractant B, the less likely he or she is toshow the predicted complementary responseto the interpersonal actions of Person A. Animportant exception occurs when the pre-

dicted complementary response to A falls atthe exact segments that define B's extremeand rigid style.

This proposition expresses the fact that amaladjusted interactant tends to respondwith his extreme and rigid style regardless ofthe segments and levels of actions presentedby Person A. Unless the person interactingwith maladjusted Interactant B presents astyle exactly complementary to B's, B willresist strongly any other prompt or bid thatwould require him or her to move to othersegments of the circle. To illustrate, Inter-actant B is defined at the extreme level ofOctant KL, that is, at K2L2 :Gullible/Merci-ful-Condoning (cf. Figure 1). If B is inter-acting with a person whose style is exactlycomplementary (i.e., at Octant NO:Sociable-Exhibitionistic), B will respond to that personwith K2L2 behaviors, and if B is interactingwith any other person, B will also respondwith K2L2 behaviors, instead of with the pre-dicted complementary response.

PROPOSITION 7. Optimal change in a per-son's interpersonal style can be effected byfacilitating an increased frequency and inten-sity of interpersonal acts from segments op-posite on the circle—by helping a person be-come more like an interactant who is semi-morphically acomplementary (behaviorallyopposite) to him or her.

This proposition, in addition to referringto general behavior change, takes interper-sonal complementarity directly into the psy-chotherapy session.

A person or client who presents a rigid andextreme maladjusted style rather ruthlesslyforces others to adopt a complementary po-sition. Over time this rigid presentation of avery restricted class of interpersonal acts rep-resents aversive stimuli for interactants, lead-ing them to avoid or escape continued trans-actions with the client. These consequencesfor the client result in a very constricted ex-perience of the world (cf. Carson, 1969; Kies-ler, 1979, 1982b; Kiesler, Bernstein, & An-chin, 1976; Leary, 1957). In order to expe-rience a greater range of experience withothers that reflects more positive conse-quences, the client needs to learn (a) to emitacts from a wider range of segments on the


circle and (b) to emit acts generally of lessintensity, that is, from the mild-moderatelevel of circle segments.

What Proposition 7 asserts is that opti-mum client change results from helping aclient to act more like a person representinghis direct opposite on the circle. To illustrate,the goal of therapeutic intervention for aclient denned at the DE:Cold-Hostile octantis an increased frequency of instances inwhich he conducts his transactions in aLM:Warm-Friendly manner. More frequentoccurrence of these acts, because they de-mand behaviors weighted exactly oppositelyon the Control and Affiliation dimensions,has the greatest likelihood of producing aspread of the client's actions away from theextreme of the DE octant.

The major reason the client has not, beenable to accomplish this goal spontaneouslyis that semimorphic-acomplementary (op-posite) responses (in this case, warm-friendlyones) trigger in him approach-avoidance ten-dencies. In other words, the client is "resis-tant" or "countercontrolling" to the changesdemanded of him to act more like his inter-personal opposite and, thus, is resistant to"getting better." Nevertheless, the major goalof interpersonal therapy is to facilitate aclient's movements toward behaviors directlyopposite to the client's denned maladjustedstyle.

PROPOSITION 8. An interactant or therapistcan exert the greatest pressure for change ina person's or client's interpersonal actions byproviding responses anticomplementary onthe circle to the person's or client's interper-sonal style.

The therapist's stance that exerts maxi-mum pressure for client change (for movingthe client toward segments opposite on thecircle) is the complement of the octant op-posite to that denning the client's style (Ben-jamin, 1974)—namely, the anticomplemen-tary octant. Another way to say this is thatthe therapist "pulls" for greatest client changewhen he or she provides, instead of the com-plementary response, the opposite of thecomplementary response—namely, the an-ticomplementary response. Continuing ourexample, for the extreme DE:Cold-Hostile

client, therapist acts exerting the greatestpressure for improvement in the client's stylewould fall at the anticomplementary octant,MN:Friendly-Sociable. As Benjamin (Note8) states, "use of the antithesis (anticomple-mentary) feature enhances the probability ofdrawing out the opposite behavior" (p. 9).

On the other hand, early use by the ther-apist of the anticomplementary responsemight result in the client's premature ter-mination, as a direct result of the strong anx-iety produced by this avoidance situation. Itseems probable, then, that effective use of theanticomplementary response can occur onlyin later stages of the therapy transaction.

Kiesler (1982b) argues that the covert com-plementary response, the impact message, isnecessarily experienced by the therapist intwo stages: a "hooked" stage in which thetherapist cannot not provide the complemen-tary response because of the superior pullingpower of the maladjusted client; and a "dis-engagement" phase in which the therapist,through various labeling and interventivemaneuvers, regains the freedom to stop thecomplementary response and move his or heractions to other quadrants of the circle. Fromthe client's perspective, it may be necessaryfor the therapist first to "hook" the client byproviding the complementary response (or,as we will see shortly, by providing asocialresponses), so that later introduction of theanticomplementary response does not elicitthe strong client anxiety probable in earliersessions.

Benjamin (Note 8) also offers some inter-esting notions regarding sequential timing ofthe anticomplementary response in therapy."With more disturbed individuals, one usu-ally cannot move directly to the antithetical(anticomplementary) point and have muchimpact" (p. 25). Instead, she offers the Shaur-ette principle to guide prior interventionsleading to eventual use of the anticomple-mentary response: "The milieu must 'con-nect' first with the patient by taking the com-plementary posture (with a constructivethrust) and then moving in a counter-clock-wise direction toward the final goal specifiedby the antithesis principle" (p. 25).

Berzins (1977) suggests similar therapeuticpossibilities in his excellent review of re-search on therapist-patient matching.


It would seem that a complementary therapist could bereadily provided. Anticomplementary pairing could beavoided since, under anticomplementary pairing con-ditions, therapeutic communication would be difficultto establish and dropouts may occur. On the other hand,given initially satisfactory (complementary) pairing con-ditions, the therapist's subsequent main task is to avoidprolonged complementary reciprocation of the patient'sinterpersonal overtures since doing so would 'confirm'the patient's rigid or constricted self-concept and littletherapeutic change could be expected, (p. 225)

As Carson (1969) earlier suggested, the gen-erally effective therapist is one who is able tomove to any quadrant of the interpersonalcircle in response to the presenting maladap-tive styles of particular patients.

DeVogue and Beck (1978) point out that"persons preferring to engage the therapistfrom a position of friendly submission shouldmake the ideal client for therapy" becausethey respond best to praise and approval.Furthermore, "these people should be mostable to assume nonaversively a compliantrole with a dominant other" (p. 234). De-Vogue and Beck go on to otfer a four-stageprocess model in which interpersonal andbehavior therapy strategies are integrated. InStage 1 "the therapist tries to avoid any stylethat would clearly fall into an intense rangeof status or affect" (p. 236). As a result of thistactic, the client begins to use his or her pre-ferred interpersonal tactic to initiate a com-fortable form of intimacy with the therapist.In Stage 2 the therapist "invites the clientdirectly into a conversation about the here-and-now relationship" (p. 236). In Stage 3the therapist "refuses to engage in the recip-rocal behavior to the client's position. In-stead, he launches an unemotional and log-ical attack against the client's position, whileproviding the client with a clear message ofhis desire to continue the relationship" (p.236). Finally, Stage 4 is "the point at whichbehavioral techniques could be optimally in-stituted" (p. 237).

PROPOSITION 9. Relative to the anticom-plementary response, an interactant or ther-apist can exert the next greatest pressure forchange in a person's or client's interpersonalstyle by emitting interpersonal acts from theclass of asocial responses, including meta-communicative feedback to the person orclient that describes his or her style and itsself-defeating consequences.

According to Beier (1966; Young & Beier,1982), whenever the therapist withholds thecustomary, preferred, or expected (i.e., com-plementary) response, the therapist respondsto the client in an asocial or disengaged way.In turn, as a direct result the client experi-ences a sense of "beneficial uncertainty," be-cause the client's preferred style did not pro-duce the expected and familiar interpersonalconsequences. Thus, the client is provided anew experience in response to his old pat-terns. If the therapist continues to providethis climate of new experience, over time theclient is obliged to discover and produce newinterpersonal acts. As examples of therapeu-tic asocial responses, Young and Beier (1982)list delay responses, reflection of content andfeeling, labeling the interaction style, and useof therapeutic paradox. Kiesler (1979, 1982a;Kiesler, Bernstein, & Anchin, 1976; Kiesler,Note 5) emphasizes throughout his interper-sonal communication psychotherapy that theasocial response of metacommunicative feed-back has interventive priority throughout thecourse of therapy.

Because the asocial response has yet to bedenned precisely relative to the interpersonalcircle, it is vital that the issue be addressedby interpersonal investigators. In my judg-ment, the asocial response relates to the in-terpersonal circle in one of three ways: (a) byfalling at the midpoint or origin of the circleand, hence, being totally neutral in regard toboth Control and Affiliation, (b) by consti-tuting a milder or more ambiguous form ofthe anticomplementary response, or (c) byfalling altogether outside the domain of thecircle on some plane denning a metalevel ofinterpersonal activity. Clarification of thisimportant issue can come only from futureresearch.

Until that time, it seems to me that anyasocial response to Person A by InteractantB requires two components: that B in no wayprovides the complementary response "bidfor" by A and that B's response has ambig-uous meaning in the sense that A cannotpinpoint at all the position on the circle fromwhich B is responding.

PROPOSITION 10. Interpersonal comple-mentarity applies primarily to naturally oc-curring, relatively unstructured interpersonal


situations. The extent to which it applies invarious structured situations or in other en-vironmental contexts remains to be deter-mined.

Several authors have argued that interper-sonal theory adopts an interactionist position(Endler & Magnusson, 1976) whereby a per-son's interpersonal acts reflect the conjointeffects of person and situation factors (Car-son, 1969; Duke & Nowicki, 1982; Kiesler,1979, 1982b; Kiesler, Bernstein, & Anchin,1976). Kiesler and his colleagues have arguedfurther that by far the most important classof situations consists of "other persons," or,more precisely, of the presenting interper-sonal styles of various interactants.

An important task for interpersonal in-vestigators is to specify classes of situationalfactors relevant (and irrelevant) to elicitationof interpersonal acts from the respective oc-tants of the interpersonal circle. For example,what kinds of situations does a DE:Cold-Hostile person tend to seek out as well as toavoid? In which contexts is his or her stylelikely to be elicited most strongly or likely tobe elicited not at all?

Duke and Nowicki (1982), addressing thepoint that most interpersonal theorists havenot specified situations beyond the generalcategory of significant others, describe an ini-tial taxonomy of situations that has guidedtheir social-learning/interpersonal researchprogram. Other situational taxonomies areappearing (e.g., Argyle, Furnham, & Gra-ham, 1981), and these or others need to beinterfaced directly with the respective octantsof the 1982 Interpersonal Circle.

Of course, the interpersonal theory artic-ulated in this article already provides a cen-tral component of any valid situationaltaxonomy by specifying exactly on the1982 Interpersonal Circle the predictedcomplementary, anticomplementary, andacomplementary octant responses for a par-ticular pair of interactants (Table 4). Com-plementary interactions evoke approach be-haviors from both participants, anticomple-mentary ones lead to avoidance or escapeactions, and acomplementary interactionsevoke a mixture of approach and avoidantresponses from both participants. Neverthe-less, it is crucial that investigators address the

interactions of these transactional patternswith other environmental or contextural fac-tors. Much is left to be done.

Until these clarifications appear, proposi-tions of complementarity seem to apply pre-cisely only to naturally occurring, relativelyunstructured transactions such as informalconversations at parties or other looselystructured social events, free-time activitiesand encounters, intimacy transactions, andopen-ended therapy interviews. The featurecommon in these situations is that minimalexpectations exist regarding socially corrector desirable responses or social role defini-tions (e.g., boss-employee, teacher-student).Essentially, an unstructured situation is onethat can appropriately elicit from interac-tants the entire range of interpersonal acts.Also, typically it is in unstructured situationsthat most interactions with significant othersoccur.

When considering situational factors, onemust keep another point in mind. That is,that the "radical trait" assumptions oftrans-situationality and transtemporality (Mischel,1968) more validly apply to maladjusted per-sons than they do to more normal (mild-moderate level) individuals. The interper-sonal definition of maladjusted behavior (ex-treme and rigid acts on the interpersonal cir-cle) indicates clearly that the actions of ab-normal individuals tend to overridedifferences in situational parameters, includ-ing different styles of interactants.

PROPOSITION 11. It is unclear how inter-personal complementarity applies over thetemporal range of continuing transactions be-tween interactants. The stages of sequentialoutcome for complementary and noncomple-mentary transactions need to be specified forall octants of the 1982 Interpersonal Circle.

Interactions represent stochastic transac-tions of varying frequency and duration, overvarying periods of lifetime. This propositionunderlines the necessity of empirically chart-ing the distinctive course that different trans-actions take. Stochastic methodologies canaccomplish this by assessing sequential de-pendencies among interactants, by revealingpatterned redundancies occurring over time.

Stochastic (sequential analysis) statisticsare summarized in Cairns (1979), Cox and

210 DONALD 3. KIESLER

Lewis (1966), Gottman (1979), Lamb, Suomi,and Stephenson (1978), Parzen (1962), andvan Hooff (1982). They include the followingspecific methodologies: (a) Markov chainanalysis (Benjamin, 1979; Hertcl, 1972; Ke-meny & Snell, 1960; Raush, 1972); (b) lagsequential analysis (Bakeman & Dabbs, 1976;Gottman & Notarius, 1978; Sackett, Note 9;(c) information theory approaches (Attneave,1959; Garner, 1962; Gottman & Notarius,1978; Losey, 1978; Raush, 1965; Steinberg,1977); (d) time-lag correlation (van Hooff,1982); and (e) the phrase structure grammarmodel (Dawkins, 1976; Hutt & Hutt, 1970;Westman, 1977).

In contrast to traditional correlational andexperimental designs, stochastic methods fo-cus on microanalysis of each unfolding ex-change between interactants over some pe-riod of time. These methods also exclusivelymeasure properties of the transacting dyad,in contrast to assessing individual behaviorwithin a dyad. Hence, they permit direct as-sessment of the propositions of complemen-tarity. Finally, stochastic methods capture allthree aspects of circular causality (Danziger,1976): feedback, redundancy, and nonsum-mativity.

Duke and Nowicki (1982) offer an initialformulation of four distinct stages of inter-action sequence: choice, beginning, deepen-ing, and termination. They speculate thateach of the four relationship phases demands"different requirements from interpersonalstyles" (p. 86). Berzins (1977) emphasizes the"need for longitudinal research that wouldclarify the manner in which participantsmove across quadrants of the InterpersonalCircle as a function of initial pairing condi-tions" (p. 237).

In sum, we have little systematic knowl-edge regarding the lawful redundancies thatcharacterize the natural sequence of distinctinterpersonal transactions. It is only by sto-chastic investigations that we can discover theimportant governing parameters as well asthe additional propositions of complemen-tarity for personality, psychopathology, andpsychotherapy.

Additional Directions for Future Research

Several expansions of interpersonal theoryare necessary for full understanding not only

of interpersonal complementarity but of hu-man behavior more generally. A brief dis-cussion of three of these issues follows.

1. Perhaps the most important future taskis to continue the creative work of Carson(1969, 1971, 1979, 1982) and Golding( 1977,1978, 1980), which attempts to specify thecognitive events that mediate interpersonalacts for persons denned at distinct octants ofthe interpersonal circle. Targeting the distor-tions in interpersonal construal style of mal-adjusted persons, Carson (1969) argues thata mutually reinforcing relationship exists be-tween the manner in which the person con-strues the social environment and his char-acteristic style of interpersonal actions.

Historically, interpersonal theory has con-centrated more on explanation of overt be-havior, with much less attention devoted tothe cognitive and other events that occur si-multaneously under the skins of individualparticipants. Many questions need answers.For example, what are the specific self-defi-nitions, cognitive assumptions, expectancies,and the like that characterize a person ex-hibiting a DE:Cold-Hostile style? What arethe specific attributes that such a person val-ues in himself or herself and in others? Orthat he or she despises? Or by which he orshe is frightened? What are the specific ra-tional and irrational self-statements that pro-pel his or her preferred self-presentations toothers? What are the cognitive styles thatshape and color his or her perceptual andother experience?

Specification of the distinctive content andstyle of a person's cognitive events is neces-sary if we are to fully understand that per-son's overt interpersonal behavior and if weare to intelligently design differential thera-peutic interventions.

2. Another important issue was referredto parenthetically in discussion of Proposi-tion 10. To clarify the effects of situationalfactors, we need to understand much moreabout significant others. Who are these per-sons? What are the defining properties of aninteractant significant to a particular person?Is a "significant other" a family member? Aspouse? A close friend? A work acquain-tance? A therapist? Some research has begunto address this central interpersonal issue(Shrauger & Schoeneman, 1979), but muchremains to be done.


3. Finally, basic aspects of nonverbal com-munication need to be integrated both withthe interpersonal circle and with interper-sonal theory more generally. For example,considerable research shows that the Controland Affiliation dimensions that underlie thecircle are also basic to nonverbal behavior(LaFrance & Mayo, 1978; Mehrabian, 1971,1972). Accordingly, the interpersonal actsdenning the 1982 Interpersonal Circle seg-ments need to include the corresponding con-trol (e.g., greater asymmetrical posture, vi-sual dominance pattern, more frequent ini-tiation of touch) and affiliation (e.g., moredirect body orientation, reciprocal gaze andtouch, closer interpersonal distance) nonver-bal behaviors.

Kiesler (1979, I982b; Kiesler, Bernstein,& Anchin, 1976; Kiesler, Note 5) provides atheoretical integration of nonverbal behaviorwith interpersonal theory. His interpersonalcommunication theory pulls together threedistinct traditions: (a) empirical research innonverbal communication (Argyle, 1975;Harper, Wiens, & Matarazzo, 1978; Harri-son, 1974; Knapp, 1978; LaFrance & Mayo,1978; Mehrabian, 1971, 1972; Siegman &Feldstein, 1978); (b) communications psy-chiatry (Bateson, 1958; Beier, 1966; Berne,1964; Goffman, 1967; Laing, 1962; Ruesch& Bateson, 1951; Watzlawick, Beavin, &Jackson, 1967; Watzlawick & Weakland,1977); and (c) interpersonal theory of per-sonality and psychopathology.

In conclusion, interpersonal theory offersexciting promise for understanding humansocial behavior and as a paradigm compre-hensive enough to encompass other theoret-ical and empirical approaches. I hope thatthis article stimulates further efforts towardfulfillment of that promise.

Reference Notes

1. LaForge, R. Using the ICL: 1976. Unpublishedmanuscript, Mill Valley, California, 1977.

2. Wiggins, J. S. Revised Interpersonal Adjective Scales.Unpublished manuscript, University of British Co-lumbia, Vancouver, Canada, 1981.

3. Kiesler, D. J. The 1982 Interpersonal Circle. Unpub-lished manuscript, Virginia Commonwealth Univer-sity, 1982.

4. Benjamin, L. S. Three levels of the SASB model(quadrant, cluster and full). (Unpublished figure.)Madison, Wis.: Intrex Interpersonal Institute, 1980.

5. Kiesler, D. J. A communications approach to modi-

fication of the "obsessive" personality: An initial for-mulation. Unpublished manuscript, Emory Univer-sity, 1973.

6. Berzins, J. I. Matching patients with therapists: Con-ceptual, empirical, and pragmatic perspectives. Paperpresented at the meeting of the Society of Psycho-therapy Research, Denver, June 1974.

7. Luborsky, L. B., & Singer, B. The fit of therapist'sbehavior into patient's negative expectations: A studyof transference-counterlransference contagion. Un-published manuscript, University of PennsylvaniaSchool of Medicine, 1974.

8. Benjamin, L. S. Validation of Structural Analysis ofSocial Rehavior (SASB). Unpublished manuscript,Wisconsin Psychiatric Institute, Madison, October1980.

9. Sackett, G. P. A nonparametric lag sequential analysisfor studying dependency among responses in behaviorobservation scoring systems. Paper presented at themeeting of the Western Psychological Association,San Francisco, April 1974.

References

Adams, H. B. "Mental illness" or interpersonal behavior?American Psychologist, 1964, 19, 191-197.

American Psychiatric Association, Diagnostic and sta-tistical manual of mental'disorders (3rd ed.). Wash-ington, D.C.: Author, 1980.

Anchin, J. C., & Kiesler, D. J. (Eds.), Handbook of in-terpersonal psychotherapy. Elmsford, N.Y.: Pergamon,1982.

Argyle, M. Bodily communication. New York: Interna-tional Universities Press, 1975.

Argyle, M., Furnham, A., & Graham, J. A. Social sit-uations. New York: Cambridge University Press, 1981.

Attneave, F. Applications of information theory to psy-chology. New York: Holt, Rinehart & Winston, 1959.

Bakeman, R., & Dabbs, I. M. Social interaction ob-served: Some approaches to the analysis of behaviorstreams. Personality and Social Psychology Bulletin,1976, 3, 335-345.

Bateson, G. Naven. Stanford, Calif.: Stanford UniversityPress, 1958.

Beery, J. W. Therapists' responses as a function of levelof therapist experience and attitude of the patient.Journal of Consulting and Clinical Psychology, 1970,34, 239-243.

Beier, E. G. The silent language of psychotherapy: Socialreinforcement of unconscious processes. Chicago: Al-dine, 1966.

Benjamin, L. S. Structural analysis of social behavior.Psychological Review, 1914,81, 392-425.

Benjamin, L. S. Structural analysis of a family in therapy.Journal of Consulting and Clinical Psychology, 1977,45, 391-406.

Benjamin, L. S. Structural analysis of differentiation fail-ure. Psychiatry, 1979, 42, 1-23. (a)

Benjamin, L. S. Use of Structural Analysis of SocialBehavior (SASB) and Markov chains to study dyadicinteractions. Journal of Abnormal Psychology, 1979,88, 303-319. (b)

Benjamin, L. S. Use of Structural Analysis of SocialBehavior (SASB) to guide intervention in psychother-apy. In J. C. Anchin & D. J. Kiesler (Eds.), Handbook

212 DONALD 3. KIESLER

of interpersonal psychotherapy. Elmsford, N.Y.: Pcr-gamon, 1982.

Berne, E. Games people play. New York: Grove Press.1964.

Berzins, J. I. Therapist-patient matching. In A. S. Gur-man & A. M. Razin (Eds.), Effective psychotherapy:A handbook of research. New York: Pergamon, 1977.

Bierman, R. Dimensions of interpersonal facilitation inpsychotherapy and child development. PsychologicalBulletin, 1969, 72, 338-352.

Buss, D. M., & Craik, K. H. The act frequency analysisof interpersonal dispositions: Aloofness, gregarious,dominance, and submissivencss. Journal of Person-ality, 1981, 49, 175-192.

Cairns, R. B. The analysis of social interaction: Methods,issues and illustrations. Hillsdale, N.J.: Erlbaum,1979.

Carson, R. C. Interaction concepts of personality. Chi-cago: Aldine, 1969.

Carson, R. C. Disordered interpersonal behavior. InW. A. Hunt (Ed.), Human behavior and its control.Cambridge, Mass.: Schenkman, 1971.

Carson, R. C. Personality and exchange in developingrelationships. In R. L. Burgess & T. L. Houston (Eds.),Social exchange in developing relationships. NewYork: Academic Press, 1979.

Carson, R. C. Self-fulfilling prophecy, maladaptive be-havior, and psychotherapy. In J. C. Anchin & D. J.Kiesler (Eds.), Handbook of interpersonal psychother-apy. Elmsford, N.Y.: Pergamon, 1982.

Celani, D. P. The complementarity hypothesis: An ex-ploratory study (Doctoral dissertation, University ofVermont, 1974). Dissertation Abstracts International,1974.

Chance, E. Content analysis of verbalizations about in-terpersonal experience. In L. A. Gottschalk and A. H.Auerbach (Eds.), Methods of research in psychotherapy.New York: Appleton-Century-Crofts, 1966.

Cox, D. R., & Lewis, P. A. W. The statistical analysisof series of events. London: Methucn, 1966.

Crowder, J. E. Relationship between therapist and clientinterpersonal behaviors and psychotherapy outcome.Journal of Counseling Psychology, 1972, 19, 68-75.

Cutler, R. L. Countertransference effects in psychother-apy. Journal oj Consulting Psychology, 1958,22, 349-356.

Danziger, K. Interpersonal communication. New York:Pergamon, 1976.

Dawkins, P. Hierarchical organization: A candidate prin-ciple for ethology. In P. G. Bateson & R. A. Hinde(Eds.), Growing pains in ethology. Cambridge, En-gland: Cambridge University Press, 1976.

DeVogue, J. T, & Beck. S. The therapist-client rela-tionship in behavior therapy. In M. Hersen, R. M.Eisler, & P. M. Miller (Eds.), Progress in behaviormodification (Vol. 6). New York: Academic Press,1978.

Dietzel, C. S., & Abeles, N. Client-therapist comple-mentarity and therapeutic outcome. Journal of Coun-seling Psychology, 1975, 22, 264-272.

Duke, M. P., & Nowicki, S. Jr. A social learning theoryanalysis of interactional theory concepts and a mul-tidimensional model of human interaction constella-tions. In J. C. Anchin & D. J. Kiesler (Eds.), Hand-

book oj interpersonal psychotherapy. Elmsford, N.Y.:Pergamon, 1982.

Edquist, M. H. Interpersonal choice and social attractionamong four interpersonal types (Doctoral dissertation,Duke University, 1973). Dissertation Abstracts Inter-national, 1973, 34, 1722B. (University Microfilms No.72-22, 977)

Endler, N. S., & Magnusson, D. (Eds.), Interactional psy-chology and personality. Washington, D.C.: Hemi-sphere Press, 1976.

Foa, U. G. Convergences in the analysis of the structureof interpersonal behavior. Psychological Review, 1961,68, 341-358.

Foa, U. G. New developments in facet design and anal-ysis. Psychological Review, 1965, 72, 262-274.

Foa, U. G. Perception of behavior in reciprocal roles:The ringex model. Psychological Monographs, 1966,80(15, Whole No. 623).

Foa, U. G., & Foa, E. B. Societal structures of the mind.Springfield, 111.: Charles C Thomas, 1974.

Garner, W. R. Uncertainty and structure as psychologicalconcepts. New York: Wiley, 1962.

Gaminsky, N. R., & Fairwell, G. F. Counselor verbalbehavior as a function of client hostility. Journal ofCounseling Psychology, 1966,13, 184-190.

Goffman, E. Interaction ritual. Garden City, N.Y.: Dou-bleday Anchor, 1967.

Golding, S. L. Individual differences in the construal ofinterpersonal interactions. In D. Magnuson & N. En-dler (Eds.), Personality at the crossroads: Current is-sues in interactional psychology. New York: Wiley,1977.

Golding, S. L. Toward a more adequate theory of per-sonality: Psychological organizing principles. In H.London & N. Hirschberg (Eds.), Personality: A newlook at metatheories. Washington, D.C.: HemispherePress, 1978.

Golding, S. L., Valone, K., & Foster, S. W. Interpersonalconstrual: An individual differences framework. In H.Nirschberg (Ed.), Multivariate methods in the socialsciences: Applications. Hillsdale, N.J.: Erlbaum, 1980.

Gottman, J. M. Detecting cyclicity in social interaction.Psychological Bulletin, 1979, 86, 338-348.

Gottman, J. M., & Notarius, C. Sequential analysis ofobservational data using Markov chains. In T. R. Kra-tochwill (Ed.), Single subject research: Strategies farevaluating change. New York: Academic Press, 1978.

Guttman, L. A new approach to factor analysis: Theradex. In P. R. Lazarsfeld (Ed.), Mathematical think-ing in the social sciences. Glencoe, 111.: Free Press,1954.

Haley, J. Strategies of psychotherapy. New York: Grune& Stratton, 1963. '

Harper, R. G., Wiens, A. N., & Matarazzo, J. D. Non-verbal communication: The state of the art. New York:Wiley, 1978.

Harrison, R. P. Beyond words: An introduction to non-verbal communication. Englewood Cliffs, N.J.: Pren-tice-Hall, 1974.

Heller, K., Myers. R., & Kline, I. Interviewer behavioras a function of standardized client roles. Journal ofConsulting Psychology, 1963,27, 117-122.

Hertel, R. K. Application of stochastic process analysisto the study of psychotherapeutic processes. Psycho-logical Bulletin, 1972, 77, 421-430.


Hutt, S. J., & Hutt, C. Direct observation and measure-ment of behavior. Springfield, 111.: Charles C Thomas,1970.

Jackson, D. D. Family interaction, family homeostasis,and some implications for conjoint family therapy. InJ. H. Masserman (Ed.), Individual and familial dy-namics. New York: Grune & Stratton, 1959.

Kemeny, J. G., & Snell, J. L. Finite Markov chains.Princeton, N.J.: Van Nostrand, 1960.

Kiesler, D. J. An interpersonal communication analysisof relationship in psychotherapy. Psychiatry, 1979, 42,299-311.

Kiesler, D. J. Confronting the client-therapist relation-ship in psychotherapy. In J. C. Anchin & D. J. Kiesler(Eds.), Handbook of interpersonal psychotherapy.Elmsford, N.Y.: Pergamon, 1982. (a)

Kiesler, D. J. Interpersonal theory for personality andpsychotherapy. In J. C. Anchin & D. J. Kiesler (Eds.),Handbook of interpersonal psychotherapy. Elmsford,N.Y.: Pergamon, 1982. (b)

Kiesler, D. J., Anchin, J. C., Perkins, M. J., Chirico, B.,Kyle, E. M., & Federman, E. J. The Impact MessageInventory. Richmond: Virginia Commonwealth Uni-versity, 1976.

Kiesler, D. J., Bernstein, A. B., & Anchin, J. C. Inter-personal communication, relationship and the behav-ior therapies. Richmond: Virginia CommonwealthUniversity, 1976.

Knapp, M. L. Nonverbal communication in human in-teraction. New York: Holt, Rinehart, & Winston,1978.

LaForge, R., & Suczek, R. F. The interpersonal dimen-sions of personality: III. An interpersonal check list.Journal of Personality, 1955, 24, 94-112.

LaFrance, M., & Mayo, C. Moving bodies: Nonverbalcommunication in social relationships. Monterey, Calif.:Brooks/Cole, 1978.

Laing, R. D. The self and others. Chicago: QuadranglePress, 1962.

Lamb, M. E., Suomi, S. J., & Stephenson, G. R. Socialinteraction analysis: Methodological issues. Madison:University of Wisconsin Press, 1978.

Leary, T. Interpersonal diagnosis of personality. NewYork: Ronald, 1957.

Lederer, W. J., & Jackson, D. D. The mirages of mar-riage. New York: Norton, 1968.

Lorr, M., & McNair, D. M. Expansion of the interper-sonal behavior circle. Journal of Personality and SocialPsychology, 1965,2, 823-830.

Lorr, M., & McNair, D. M. The Interpersonal BehaviorInventory, Form 4. Washington, D.C.: Catholic Uni-versity of America, 1967.

Losey, G. S. Information theory and communication. InP. W. Colgan (Ed.), Quantitative ethology. New York:Wiley, 1978.

Malone, T. W. System simulation: Computer simulationof two-person interactions. Behavioral Science, 1975,20, 260-267.

McLemore, C. W, & Benjamin, L. S. Whatever hap-pened to interpersonal diagnosis? A psychosocial al-ternative to DSM-III. American Psychologist, 1979,34, 17-34.

McLemore, C. W., & Hart, P. P. Relational psychother-apy: The clinical facilitation of intimacy. In J. C. An-

chin & D. J. Kiesler (Eds.), Handbook of interpersonalpsychotherapy. Elmsford, N.Y.: Pergamon, 1982.

Mehrabian, A. Silent messages. New York: Wadsworth,1971.

Mehrabian, A. Nonverbal communication. Chicago: Al-dine, 1972.

Mischel, W. Personality and assessment. New York:Wiley, 1968.

Mueller, W. J. Patterns of behavior and their reciprocalimpact in the family and in psychotherapy. Journalof Counseling Psychology Monographs, 1969, /6(No.2, Part 2).

Mueller, W. J., & Dilling, C. A. Therapist-client interviewbehavior and personality characteristics of therapists.Journal of Projective Techniques and Personality As-sessment, 1968, 32, 281-288.

Mueller, W. J., & Dilling, C. A. Studying interpersonalthemes in psychotherapy research. Journal of Coun-seling Psychology, 1969, 16, 50-58.

Oden, T. C. TAG: Transactional awareness game. NewYork: Harper & Row, 1976.

Parzen, E. Stochastic processes. San Francisco: Holden-Day, 1962.

Perkins, M. J., Kiesler, D. J., Anchin, J. C., Chirico,B. M., Kyle, E. M., & Federman, E. J. The ImpactMessage Inventory: A new measure of relationship incounseling/psychotherapy and other dyads. Journal ofCounseling Psychology, 1979, 26, 363-367.

Raush, H. L. Interaction sequences. Journal of Person-ality and Social Psychology, 1965, 2, 487-499.

Raush, H. L. Process and change: A Markov model forinteraction. Family Process, 1972, 11, 275-298.

Raush, H. L., Dittman, A. T, & Taylor, T. J. The inter-personal behavior of children in residential treatment.Journal of Abnormal and Social Psychology, 1959, 58,9-27.

Raush, H. L., Farbman, L., & Llewellyn, L. G. Person,setting and change in social interaction: II. A normal-control study. Human Relations, 1960, 13, 305-333.

Rice, P. L. K. The modification of interpersonal roles(Doctoral dissertation, West Virginia University, 1969).Dissertation Abstracts International, 1970, 30, 4797B.(University Microfilm No. 70-6329)

Ruesch, J., & Bateson, G. Communication: The socialmatrix of psychiatry. New York: Norton, 1951.

Schaefer, E. S. A circumplex model for maternal behav-ior. Journal of Abnormal and Social Psychology, 1959,59. 226-235.

Schutz, W. C. FIRO: A three-dimensional theory of in-terpersonal behavior. New York: Holt, Rinehart &Winston, 1958.

Shannon, J., & Guerney, B. Jr. Interpersonal effects ofinterpersonal behavior. Journal of Personality and So-cial Psychology, 1973,26, 142-150.

Shrauger, J. S., & Schoenemann, J. J. Symbolic inter-actionist view of self-concept: Through the lookingglass darkly. Psychological Bulletin, 1979, 86, 549-573.

Siegman, A. W., & Feldstein, S. (Eds.). Nonverbal be-havior and communication. Hillsdale, N.J.: Erlbaum,1978.

Sluzki, C. E., & Beavin, J. Symmetry and complemen-tarity: An operational definition and a typology ofdyads. In P. Watzlawick & J. H. Weakland (Eds.), Theinteractional view. New York: Norton, 1977.


Smelser, W. T. Dominance as a factor in achievementand perception in cooperative problem solving inter-actions. Journal of Abnormal and Social Psychology,1961, 62, 535-542.

Steinberg, J. B. Information theory as an ethological tool.In B. A. Hazlett (Ed.), Quantitative methods in thestudy of animal behavior. New York: Academic Press,1977.

Sullivan, H. S. The interpersonal theory of psychiatry.New York: Norton, 1953.

Swenson, C. H. Psychotherapy as a special case of dyadicinteraction: Some suggestions for theory and research.Psychotherapy: Theory, Research and Practice, 1967,4. 7-13.

van Hooif, J. A. R. A. M. Categories and sequences ofbehavior. In K. L. Scherer & P. Ekman (Eds.), Hand-book of methods in nonverbal behavior research. NewYork: Cambridge University Press, 1982.

Watzlawick, P., Beavin, J. H., & Jackson, D. D. Prag-matics of human communication. New York: Norton,1977.

Watzlawick, P., & Weakland, J. H. (Eds.), The interac-tional view. New York: Norton, 1977.

Westtnan, R. S. Environmental languages and the func-tional bases of behavior. In B. A. Hazlett (Ed.), Quan-titative methods in the study of animal behavior. NewYork: Academic Press, 1977.

Wiggins, J. S. A psychological taxonomy of trait-descrip-tive terms: The interpersonal domain. Journal of Per-sonality and Social Psychology, 1979, 37, 395-412.

Wiggins, J. S. Circumplex models of interpersonal be-havior in clinical psychology. In P. C. Kendall & J. N.Butcher (Eds.), Handbook of research methods in clin-ical psychology. New York: Wiley, 1982.

Winch, R. F. Mate-selection: A study of complementaryneeds. New York: Harper & Row, 1958.

Young, D. M., & Beier, E. G. Being asocial in socialplaces: Giving the client a new experience. In J. C.Anchin & D. J. Kiesler (Eds.), Handbook of interper-sonal psychotherapy. Elmsford, N.Y.: Pergamon, 1982.

Received October 4, 1982Revision received January 28, 1983 •

Third Edition of the Publication Manual

APA has just published the third edition of the Publication Manual. This newedition replaces the 1974 second edition of the Manual. The new Manual updatesAPA policies and procedures and incorporates changes in editorial style and practicesince 1974. It amplifies and refines some parts of the second edition, reorganizesother parts, and presents new material. (See the March issue of the American Psy-chologist for more on the third edition.)

All manuscripts to be published in the 1984 volumes of APA's journals will becopy edited according to the third edition of the Manual. Therefore, manuscriptsbeing prepared now should be prepared according to the third edition. Beginningin 1984, submitted manuscripts that depart significantly from third edition style willbe returned to authors for correction.

The third edition of the Publication Manual is available for $ 12 for members ofAPA and $15 for nonmembers. Orders of $25 or less must be prepaid. A charge of$1.50 per order is required for shipping and handling. To order the third edition,write to the Order Department, APA, 1400 N. Uhle Street, Arlington, VA 22201.

本文献由“学霸图书馆-文献云下载”收集自网络，仅供学习交流使用。

学霸图书馆（www.xuebalib.com）是一个“整合众多图书馆数据库资源，

提供一站式文献检索和下载服务”的24 小时在线不限IP

图书馆。

图书馆致力于便利、促进学习与科研，提供最强文献下载服务。

图书馆导航：

图书馆首页文献云下载图书馆入口外文数据库大全疑难文献辅助工具

http://www.xuebalib.com/cloud/

http://www.xuebalib.com/

http://www.xuebalib.com/cloud/


http://www.xuebalib.com/vip.html

http://www.xuebalib.com/db.php

http://www.xuebalib.com/zixun/2014-08-15/44.html


the 1982 interpersonal circle: a taxonomy for ... · described corresponding overt interpersonal...

Documents