You are on page 1of 6

Psychotherapy Research

ISSN: 1050-3307 (Print) 1468-4381 (Online) Journal homepage: http://www.tandfonline.com/loi/tpsr20

The time has come: The real relationship in


psychotherapy research

Charles J. Gelso

To cite this article: Charles J. Gelso (2009) The time has come: The real relationship
in psychotherapy research, Psychotherapy Research, 19:3, 278-282, DOI:
10.1080/10503300902777155

To link to this article: http://dx.doi.org/10.1080/10503300902777155

Published online: 06 Jul 2009.

Submit your article to this journal

Article views: 246

View related articles

Citing articles: 6 View citing articles

Full Terms & Conditions of access and use can be found at


http://www.tandfonline.com/action/journalInformation?journalCode=tpsr20

Download by: [Universidad de los Andes] Date: 18 May 2017, At: 06:29
Psychotherapy Research, May 2009; 19(3): 278282

REAL RELATIONSHIP

The time has come: The real relationship in psychotherapy research

CHARLES J. GELSO

Department of Psychology, University of Maryland, College Park, Maryland, USA


(Received 6 January 2009; revised 22 January 2009; accepted 24 January 2009)

Let me begin by expressing my appreciation to Drs. just have to settle for the likelihood that there is no
Leigh McCullough, Robert Hatcher, and Adam widely appealing term for what I refer to as the real
Horvath for preparing such high-quality reactions or personal relationship. I suspect the field can live
to my article. All three reactions, it seemed to me, with this, while staying mindful of the idea that the
were thoughtful, respectful, and useful. I could not substance or processes the term depicts are of far
have asked for more. It was clear that these three greater significance than the term itself.
scholars each had some views that were similar to
mine and some that were different. Regardless of the
Some Theoretical Matters
degree to which we differed, I found each comment
to be intellectually exciting, and each gave me new Must the Good Real Relationship Be Positive?
ideas to explore. In this rejoinder, rather than reply
McCulloughs (2009) observations about real rela-
to each and every reaction, I shall focus mostly on a
tionships and the three distinctions that need further
few points of contention that seem central and seek
attention seem very important. That is, the distinc-
to provide clarifications of what may be misunder- tions between realness versus genuineness, real
standings. relationship versus therapeutic alliance, and real
relationship versus transference are all crucial to
the clarification of what constitutes a real relation-
Semantics that Will Not Go Away
ship. These distinctions are worthy of a careful
Before replying to substantive theoretical and mea- reading by anyone interested in the real relationship
surement matters, I do want to reiterate that the construct. Despite my sense that McCullough and
term real relationship does carry its share of I seem to be of a like mind about the therapeutic
problems. I have considered alternative terms over relationship, there is one clinical/theoretical differ-
the years. None fully fit the processes to which the ence between our views that may fruitfully be high-
term applies without carrying controversial excess lighted. McCullough takes me to task for suggesting
baggage, although, as I have also noted, I have been that positively valenced ratings are needed if the real
adding the term personal relationship more often relationship is to be considered strong. Instead, she
in recent times. This term, too, has its share of believes that the real relationship allows the freedom
problems, as Leigh McCullough highlights. Inter- (presumably for therapist as well as patient) to
estingly, whereas McCullough preferred the term confront conflictual and negative feelings toward
real relationship and believed that personal one another, which, in turn, may strengthen the real
relationship had undesirable connotations, Adam relationship.
Horvath (2009) strongly disapproved of real rela- I actually am in close agreement with these
tionship, preferring the term personal relation- assertions. In fact, in one of the studies that was
ship. To wit, in searching for a better term, Frank part of the Maryland Psychotherapy Relationship
(2005) mentioned the new relationship, which to Research Program (Woodhouse, Schlosser, Crook,
my mind is even more problematic. I think we shall Ligiero, & Gelso, 2003), we found that clients who

Correspondence concerning this article should be addressed to Charles J. Gelso, Department of Psychology, Biology-Psychology Building,
University of Maryland, College Park, MD 20742-4411, USA. E-mail: gelso@psyc.umd.edu

ISSN 1050-3307 print/ISSN 1468-4381 online # 2009 Society for Psychotherapy Research
DOI: 10.1080/10503300902777155
The time has come 279

were more securely attached to their therapists manner of studying constructs will be fruitful,
(indicative of a strong real relationship) exhibited I believe that the exclusive use of that approach,
greater overall transference and negative transfer- and the elimination of the component point of view,
ence toward their therapists, as judged by these will itself narrow and limit our understanding. This
therapists. To my mind, these findings imply that a is so because, although the real relationship, for
secure attachment to ones therapist, and by im- example, may be part of everything that happens in
plication of strong real relationship, gives patients therapy, including the enactment of techniques,
the freedom to explore negative feelings toward the there are times when one or another component
therapist. Also, in a qualitative study (Gelso, Hill, (e.g., relationship or technique; real relationship,
Mohr, Rochlen, & Zack, 1999), therapists who working alliance, transference, and countertransfer-
described cases of successful, long-term psychody- ence) will be clearly the most prominent phenom-
namic psychotherapy typically viewed a positive real enon that is being exhibited and, further, at times
relationship in these cases as creating a secure base, studying mutually exclusive components is quite
which allowed patients negative transference to useful. As an example, although Hatcher appears
emerge and be worked through. At the same time, to think that it is not useful to study the importance
it seems to me that in order for the real relationship to outcome of the working alliance versus technique,
to be strong, it must be largely positively valenced or the real relationship versus the alliance, or
from both the therapists and the patients perspec- transference versus nontransference, I could see
tives. A positively valenced, strong real relationship where each of these comparisons might be extremely
will likely provide the patient with a secure base from enlightening. Would such a component approach be
which negative feelings (both transferential and sufficient? I agree with Hatcher that it would not.
realistic) may be dealt with and resolved. Should it be done away with in favor of what might
I do tend to agree with McCullough that the real be termed his all-in-everything approach? In my
relationship may be even more important, or at least view, it should not. Instead, here we have a great
more prominent, in brief therapy than long-term situation for theoretical and methodological trian-
treatment (see Gelso & Hayes, 1998, pp. 125126). gulation. We can theorize about and study phenom-
However, its import in longer term treatment should ena from multiple perspectives, with, I believe, a
not be underestimated. For example, the Gelso et al. superior yield.
qualitative study hints at the major significance of
the real relationship in long-term dynamic therapy in
terms of the resolution of transference. What Is Real Anyway?
Adam Horvath (2009) offers a number of interesting
and potentially fruitful ideas about the therapeutic
What Is the Best Unit of Analysis?
relationship. For example, his conception of layers of
In a theoretical statement that has implications for constructs (ranging from feelings at the first level
research, Hatcher (2009) explores in some detail the and relationship processes at the highest level of
extent to which the real relationship ought to be abstraction) is intriguing, and I could easily see it
framed as a way of looking at what happens in each developing into a general structure for understand-
and every therapy transaction, in contrast to a ing the therapeutic relationship, a structure that
phenomenon that shows itself at certain times in would take the field beyond the excessively global
concrete aspects of the therapy exchange. Hatcher conceptualizations of the therapeutic relationship
strongly and clearly favors the former manner of that both he and I lament. Of course, the conception
formulating the real relationship. The latter, which would need much refinement and development
he refers to as the component point of view, before it might have such an impact.
narrows and limits our potential understanding Regarding the real relationship, Horvaths remarks
because any given activity, attitude, etc. in therapy represent the most fundamental challenge to a
serves multiple purposes, and thus cannot in general theory of the real relationship in psychotherapy as
belong in an exclusive category (p. 269). In other I have formulated it. Specifically, Horvath is critical
words, the component approach is problematic of the concept of realism within the real relationship.
because it measures the strength of a given compo- He sees my solution to the question of what is real
nent at a particular time or on the whole, whereas as involving the notion of fit. In other words,
Hatcher believes it best to analyze, for example, the whether ones experience and perception of the other
real relationship element in every transaction. is realistic depends on how closely that perception
Although I agree with Hatcher that there are real matches the self-perception of the other. Although
relationship (and transference) aspects in virtually all I can understand why Horvath thinks that this is
transactions, and further agree that his suggested what I think because of a couple of examples
280 C. J. Gelso

I offered in the main article, this is actually not my therapies, although how it is handled depends on the
solution to the age-old ontological question of what particular theory and therapistpatient pair. More
constitutes real or reality. In my view, there is generally, constructs originating in one theoretical
indeed a self, a mind, a psychic reality, and a basic system may be extrapolated to another theory if it
personality of the patient (and the therapist, too) makes sense that these constructs apply to the other
that transcends perceptions held by the other. Thus, theories. To the point of my main article, although
for example, the person exists as an entity and is far the real relationship is a concept originating in
more than merely the result of coconstructions by psychoanalysis (as is the working alliance, by the
patient and therapist in their work together, as Eagle way), considering how it has been defined, it is hard
(2003) has aptly argued. There is indeed a reality of to imagine this component of the overall relationship
the patient (and the therapist), and a strong real not existing to one degree or another in all ther-
relationship involves each participant grasping or apeutic and all human relationships.
sensing that reality. (Of course, this too is a matter of
degree.) At the same time, in our measurement
approach, we do rely on patient and therapist reports Measuring the Real Relationship
(of self, other, and relationship). That is why I Item Generation and Refinement
have termed our approach constructive realism.
There is a reality of the patient and the therapist, and Perhaps based on my brief summaries of the projects
the participants perceptions of self and other are that have been completed by our research teams, it
extremely important in attempting to capture that seems to me that the comments evidenced a few
reality. Can a given participants perceptions be misunderstandings of our instrument development
inaccurate? They certainly can and indeed always research. For example, McCullough states that her
are to one degree or another. We need to stay main concern about the instrument development
mindful of this fact when we measure realism work that my collaborators and I have done is that
through patient and therapist reports (or reports of we bypassed the standard approach to measure
others, for that matter). development. Although our measures are theory
driven rather than empirical, such as the Minnesota
Multiphasic Personality Inventory, we actually did
Transporting Constructs Across Theories follow a fairly standard path for theory-driven
Another very significant criticism offered by Horvath measures. For example, for the therapist version of
is that because the constructs of transference and real the Real Relationship Inventory (RRI; Gelso et al.,
relationship originate from psychoanalytic theory, 2005), we generated 130 items aimed at capturing
and because transference in particular is addressed theoretically determined subscales of Realism and
much more frequently in psychoanalytic than other Genuineness and within these subscales, magnitude
treatments, the concepts of transference and what and valence. Further, we wrote items focused on the
Horvath views as its twin, the real relationship, therapist, the client, and the relationship within
cannot be framed as general elements of all ther- genuineness, realism, magnitude, and valence. We
apeutic relationships. To the contrary, in my view, then used standard item-reduction methods (e.g.,
although transference is, of course, not central to back-translation, reduction in alpha when item
nondynamic theories, it is still legitimate to theorize deleted, inspection for redundancy, and itemtotal
that transference is part of all treatments (and all correlations) to pare down the measure to the 12
relationships), whether dealt with directly or not by best items within the Realism and Genuineness
the therapist in any given treatment. The originators subscales, with at least one item representing mag-
of key nonanalytic theories such as gestalt, client- nitude and valence related to self, client, and
centered, rational-emotive, and cognitive therapies relationship for the two subscales. We conducted
all believed that transference does exist, and research confirmatory factor analysis, and we assessed inter-
evidence supports such views, even, for example, in nal consistency and both convergent and discrimi-
behavioral therapies (see Gelso & Hayes, 1998, for nant validity. As you can see, this gets pretty
discussions of transference in nonanalytic therapies), complex, and in writing the original article, I hoped
although it is not centrally worked with as such in to avoid such complexity for fear of both putting the
these treatments, unless it becomes a problem. Of reader into a state of terminal numbness and making
course, this is a matter of degree, and therapies the article much longer than it ought to be.
espousing the importance of transference are more Hatcher makes several valuable suggestions about
likely to infer more transference and bring more of measurement. For example, he suggests the kinds of
it to the surface. Moreover, I am suggesting items that might better fit his preference for studying
that transference is usefully seen as existing in all the real relationship as it exists in all experiences and
The time has come 281

transactions in therapy. Although it was unclear to do so, evidence is needed of nonoverlapping variance
me how his exemplified items would capture his with proximal relational variables such as positive
ideas better than our existing items, I could see transference, secure attachment, and alliance. In
where a measure based on his views would be response, we have indeed accumulated evidence of
potentially quite useful. such nonoverlapping variance in the four studies on
which I report, except for the correlation of the RRI-
Client with the Working Alliance InventoryBond
Do Realism and Genuineness Predict subscale. As Horvath aptly notes, when we correct
Differentially? for unreliability of coefficients, it looks like the RRI-
Both Hatcher and McCullough were interested in Client is essentially the same thing as the Bond
the possible differences in how the Realism and subscale of the Working Alliance Inventory (short
Genuineness subscales of the RRI might correlate version). In inspecting the four-item subscale, it
with other important variables, despite the very high looks like three of the items tap to a substantial
intercorrelation between realism and genuineness. extent the personal or real relationship and are
Hatcher was puzzled as to why I did not report the different from the working collaboration that seems
subscale correlations with other measures more theoretically to define a working alliance. In this
often, and McCullough expressed the view that sense, the Bond subscale of the Working Alliance
further research might reveal differences between Inventory appears to be a mixture of working
the Realism and Genuineness subscales. I am in full alliance and real relationship, whereas the RRI may
agreement that the subscale correlations with other be a cleaner measure of real relationship.
variables should be reported and that we should Although Horvath believes that the concept of
maintain an interest in how realism and genuineness genuineness has greater appeal than does realism, he
may differentially relate to other variables. In fact, in suggests that the research community might be
three of the four studies I have summarized, my better served by using the Barrett-Lennard Relation-
collaborators and I did indeed conduct separate ship Inventory (BLRI; Barrett-Lennard, 1962) than
analyses according to the two subscales. Because the RRI. The problem with this suggestion is that the
the two subscales were so highly correlated (rs ].75 BLRI taps only perceptions of the therapists genu-
in three studies) and, again, to avert excessive detail ineness. Although this is important to tap, and the
in the article on which McCullough, Hatcher, and BLRI is an effective measure of it, I believe the real
Horvath comment, I presented primarily our find- or personal relationship is better viewed as a
ings about the total score. However, despite their bipersonal construct, which includes assessments of
high intercorrelation, we have found some intriguing both therapist and client genuineness as well as the
differences in how the Realism and Genuineness genuineness of their relationship. Thus, measures
subscales of the RRI behave. For example, in Gelso completed by both therapist and client, with items in
et al. (2005), we found that therapists responses to each measure pertaining to both therapist and client
the RRIGenuineness subscale were more strongly (and their relationship), are needed. The RRI
correlated with working alliance (r .55) than were measures do reflect this bipersonal perspective.
responses to the RRI-Realism subscale (r .32), Finally, in contrast to Horvath, I do believe that
t(89) 5.13, p B.01. At the same time, in another we very clearly have accumulated sound reliability
one of our studies (Marmarosh et al., in press), we and validity data on the client and therapist RRI
found that on the therapist measure of the RRI, measures, such that research using these measures is
Realism scores predicted treatment outcome, viable and will advance knowledge. Thus, as indi-
whereas Genuineness scores did not. Given the cated in Gelso (2009), research using the current
high intercorrelation of realism and genuineness, it measures ought to be done simultaneously with
is entirely possible that these differences may have research aimed at measure refinement, such as, for
been chance findings. However, I agree with McCul- example, the research Hatcher and Gillaspy (2006)
lough that the question of whether the two subscales conducted on the Working Alliance Inventory
should be combined merits the tolerance of ambi- (Horvath & Greenberg, 1989). As for the other
guity for some time longer (p. 266). constructs Horvath believes should be considered as
components of a personal relationship (e.g., respon-
siveness, warmth, and sincerity), these all seem like
Validity and the Choice of Measures
interesting candidates. However, for these candi-
Horvath makes many good points about instrument dates to be accepted, someone needs to develop a
development. For example, we clearly need to different theory of the real relationship than I have
study further the discriminant validity of the RRI- presented and construct a different measure or two.
Therapist and the RRI-Client. He suggests that to Doing so would certainly move the science of the
282 C. J. Gelso

real or personal relationship in psychotherapy for- Bordin, E. S. (1979). The generalizability of the psychoanalytic
ward. concept of the working alliance. Psychotherapy: Theory,
Research, and Practice, 16, 252260.
Eagle, M. N. (2003). The postmodern turn in psychoanalysis: A
Conclusion critique. Psychoanalytic Psychology, 20, 411424.
Frank, K. A. (2005). Toward conceptualizing the personal
Where are we with respect to the real relationship? In relationship in therapeutic action: Beyond the real relation-
the words of Lewis Carrolls Walrus, The time has ship. Psychoanalytic Perspectives, 3(1), 1556.
come. However, although the time has come to Gelso, C. J. (2009). The real relationship in a post-modern world:
Theoretical and empirical explorations. Psychotherapy Research,
talk of many things about the real relationship, at
19(3), 253264.
this point what will be especially helpful is actual Gelso, C. J., & Carter, J. A. (1985). The relationship in counseling
empirical research aimed at addressing the many and psychotherapy: Components, consequences, and theore-
questions that have been pointed to by McCullough, tical antecedents. The Counseling Psychologist, 13, 155243.
Hatcher, Horvath, and me. Vigorous empirical study Gelso, C. J., & Carter, J. A. (1994). Components of the
of the real relationship will itself lead to more refined psychotherapy relationship: Their interaction and unfolding
during treatment. Journal of Counseling Psychology, 41, 296306.
theory. We all understand that the construct that has
Gelso, C. J., & Hayes, J. A. (1998). The psychotherapy relationship:
been termed the real or personal relationship is Theory, research, and practice. WileyNew York.
maddeningly complex. So are transference, techni- Gelso, C. J., Hill, C. E., Mohr, J. J., Rochlen, A. B., & Zack, J.
que, and working alliance for that matter. The latter (1999). Describing the face of transference: Psychodynamic
construct may not seem so complex now because therapists recollections about transference in cases of success-
Horvath and Greenberg (1989) did the hard work of ful long-term therapy. Journal of Counseling Psychology, 46,
257267.
developing a reasonable, if fallible, measure based on
Gelso, C. J., Kelley, F. A., Fuertes, J. N., Marmarosh, C., Holmes,
a highly fallible but heuristically rich theory of S. E., Costa, C., & Hancock, G. R. (2005). Measuring the real
working alliance offered by Bordin (1979). It seems relationship in psychotherapy: Initial validation of the therapist
to me that we are at a point, unlike when Jean Carter form. Journal of Counseling Psychology, 52, 640649.
and I offered our initial theoretical propositions Hatcher, R. L. (2009). Considering the real relationship: Reaction
about the therapeutic relationship (Gelso & Carter, to Gelsos The real relationship in a post-modern world:
1985, 1994), where research on real relationship is Theoretical and empirical explorations. Psychotherapy
Research, 19(3), 269272.
quite practicable. We have the psychometrically Hatcher, R. L., & Gillaspy, J. A. (2006). Development and
sound initial measures and an array of good ques- validation of a revised version of the Working Alliance
tions based on theory, practice, and beginning Inventory. Psychotherapy Research, 16, 1225.
research efforts; and there has been a growing Horvath, A. O. (2009). How real is the real relationship?
awareness, as implied in Horvaths comment, that Psychotherapy Research, 19(3), 273277.
it is time to move beyond a highly global concept of Horvath, A. O., & Greenberg, L. S. (1989). The development and
validation of the Working Alliance Inventory. Journal of
the therapeutic relationship and beyond the view
Counseling Psychology, 36, 223233.
that the therapeutic relationship is the equivalent of Marmarosh, C. L., Gelso, C. J., Markin, R. D., Majors, R.,
the working alliance. Vigorous empirical study of the Mallery, C., & Choi, J. (in press). The real relationship in
real relationship will serve to refine theory, allow us psychotherapy: Relationships to adult attachments, working
to learn more about the place of real relationship in alliance, transference, and therapy outcomes. Journal of Coun-
the overall therapeutic relationship, and ultimately seling Psychology, 19(3).
McCullough, L. (2009). The challenge of distinguishing figure
help facilitate effective real relationships and more
from ground: A reaction to Gelsos work on the real relation-
efficacious psychotherapy. ship. Psychotherapy Research, 19(3), 265268.
Woodhouse, S. S., Schlosser, L. Z., Crook, R. E., Ligiero, D. P., &
References Gelso, C. J. (2003). Client attachment to therapist: Relations to
transference and client recollections of parental caregiving.
Barrett-Lennard, G. T. (1962). Dimensions of therapists
Journal of Counseling Psychology, 50, 395408.
response as causal factors in therapeutic change. Psychological
Monographs, 76(Whole No. 43).

You might also like