Tuesday, February 09, 2010


The relationship between emotionally expressive writing (EW) as a coping mechanism and specific brain functions has never been fully detailed. Perhaps due to EW‘s longstanding favor within the field of positive psychology, and even more so at the forefront of popular psychology, all proposed neuroanatomical effects have never been outlined in a single review. This paper examines the evidence from learning skills and memory augmenting studies, as well as long-term cognitive longevity findings, while it discounts less quantifiable reports and ostensible benefits, in order to determine whether writing affords any striking cognitive gains as a therapeutic strategy. Among the reviewed findings are gains in working memory capacity (WMC) and pedagogical enhancement strategies. It is proposed here that mainstream media accounts of some purported cognitive benefits of EW are misleading, and therefore a clarification is in order: the mere act of EW itself matters less qualitatively than the individual actor’s ongoing pattern of interpretation of negative or positive life events—but EW does matter.

Though it had long been an area of interest for a segment of investigators, until recent years, expressive writing (EW) had not gained the attention of mainstream audiences. No one has done more to change the trajectory than Dr. David A. Snowdon, whose Nun Study has traced the cognitive declines of populations of Notre Dame nuns, since 1986. Snowdon‘s work has yielded many landmark findings, among them a correlation between Alzheimer pathology (AD) and strokes (Snowdon, Greiner, Mortimer, 1997), and a 100 percent success rate using artificial neural networks (ANNs) to diagnose AD (Grossi & Buscema, 2007). Perhaps the most talked about finding in mainstream press is Danner and Snowdon‘s metanalysis of a cache of decades-old autobiographies, written by prospective nuns prior to their formal entrance into the order; their study revealed a striking positive correlation between early-life emotional expression and late-life cognitive potency (2001). The nuns who composed the most idea-dense, grammatically complex sentences amidst their autobiographies experienced, life-wise, the least overall cognitive declines (Snowdon & Greiner, 1999). This was translated by the mainstream media into medically validated findings that becoming a teacher equaled less cognitive decline (nuns predominately taught), as did greater amounts of positive self-talk. Even TV hosts joined in, espousing the merits of so-called ―gratitude journals.‖
Wherever one comes across on this issue, whether with Oprah Winfrey and her advice that people must exercise the past-time of appreciation writing every day, or whether you think Snowdon‘s widely publicized work (Manning, 2001; Lemonick, 2001; Weiss, 1997; Belluck, 2001) has simply revealed an interesting personality finding, a review of the existing literature is warranted. With that, it is proposed here that a clarification is needed among mainstream media accounts of related findings: while EW offers a successful avenue for emotional release, it does not, in itself, provide any striking and long-term cognitive benefits, outside of some evidence from motor skills & learning literature, which is examined below. The clearest cognitive benefits of EW seem to exist within the narrow spectrum of self-interpretation of emotional content. For example, an individual may compose an emotionally positive essay; there are a myriad of reasons why, when prompted, someone might do this, other than that it accurately reflects an internal cognitive state. The benefits of EW and its parent concepts, summarized below, appear to be experienced not merely via the act of writing, but via a regimen, wherein an actor makes clear cognitive connections between positivity and ongoing events in one‘s own life.

Memory Augmentation

Whether EW offers anything quantifiable to the field of neuroscience is a question perhaps best measured by experiments structured around memory tasks. A recent Japanese study examined a proposed relationship between working memory capacity (WMC) and EW (Yogo & Fujihara, 2008). This finding had already been substantiated in earlier literature (see below); however, the investigators sought to determine whether the finding itself would translate to a collectivist culture, one more decorous than its westernized counterparts. Regarding the very use of EW as a diagnostic tool/stress coping mechanism among non-native study participants (originally from collectivist cultures) enrolled at an urban U.S. university, the results of Tavakoli and colleagues (2009) clearly support Yogo.
In the Yogo study, 104 undergraduate students were assigned to one of three conditions: write about a personal traumatic event (TE), write a prospective account of your best possible self (BPS), or write about a trivial topic for 20 minutes. Using word list memorization tasks to measure WMC in post-EW sessions, researchers found an association between immediate mood improvement and the BPS condition, whereas the TE condition indicated sustainable memory gains across subsequent weeks.
Likewise, in the parent study, Klein and Boals found that TE- EW freed the participant‘s WMC; the authors theorized this occurred via a specific reduction of intrusive thoughts about stressful experiences (2001). The structure of this study (two versions were performed to clarify ambiguous data) was generally the same as the Japanese study, with the exception of a positive event (PE) writing group, rather than a BPS group, and a longer experiment time-frame.
As was the case in the Japanese study, PE- EW exhibited no effect on WMC. Why? Extrapolating from existing literature, Klein and Boals note that much of one‘s day to day cognitive capacity is consumed by thoughts of traumatic and/or stressful events. Expressive writing frees the actor from such a cognitive load, allowing other tasks to take precedence, and negative health-wise benefits not to take hold (2001).
The authors found that, as measured by performance on word list memorization tasks, TE- EW not only improved WMC for a period of at least 7 weeks, but the quality of the writing in the event (PE or TE) versus trivial tasks was also noticeably different. In both event writing groups, greater narrative coherence was exemplified by higher levels of cause and insight words. Where the disparity was found, however, appears to be in the mental processes for PE versus TE narrative reconstruction. Klein asks whether positive occurrences simply exert less of a cognitive impact, or if there is something yet quantifiably unknown about writing about negative experiences.
Lastly, Klein and Boals examined whether as WMC improved, so did Grade Point Average (GPA). Existing literature had already found that EW benefitted GPA. Because WMC could now be seen as benefitting from EW, it was inferred that WMC improvement would equal GPA improvement. This was clearly borne out by the data as well.
To reiterate, mere emotive content had no lasting effect on WMC. Study participants in the PE-EW groups experienced no quantifiable gains, whereas those in NE-EW groups did. Taken with the above, cause and insight words, plus the mental exertions required to recreate an NE versus a PE account, it appears that participants‘ personal insight regarding their writing is more important than mere emotional expression. Klein and Boals point to non-writing studies that have borne out these effects (Lutgendorf, Antoni, Kumar & Schneiderman, 1994; Kelley, Lumley & Leisen, 1997).
This study is not the first to posit that the narrative reconstruction element of EW is of paramount importance. Klein and Boals specifically note that the creation of mental models, which package TE s and other generic stressors into something more manageable, may be what allows for the gains seen. However, Pennebaker suggests more generally that the introductory step of emotional expression, with regard to a traumatic memory, can begin a cascade of cognitive processes which culminates in physiological changes (1989, 1993). What Klein and Boals call ―models‖ or ―packages,‖ Pennebaker and colleagues call ―linguistic structural promoters,‖ assimilators that allow the actor to lessen stressful cognitions with regard to TEs. This is the transitional step which soon equals overall mood improvement (Pennebaker, Mayne & Francis, 1997).
According to Foa and others, an important therapeutic technique in the treatment of traumatic memories is continued processing, an end result being improved memory organization; the disorganization of memory is proposed to be the major psychological stressor (Foa, Steketee & Rothbaum, 1989; Pennebaker, et. al., 1997). There are a myriad of methods to this end, and traditional art therapy may prove significant in the future. However, for our purposes, EW as realized in a fully integrated narrative form shows much promise as a potential therapeutic technique (Foa & Kozak, 1986; van der Hart, Steele, Boon & Brown, 1993).

Learning Augmentation

If memory is influenced by EW, it can be inferred that memory‘s subsequent action step, learning, is also quantifiably influenced by, if not EW specifically, then the parent concept of simple structural writing. In 2008, Longcamp and colleagues studied participants‘ novel character recognition, and found significant differences in brain neuroanatomy among those in either written trials or typed trials. Researchers created unique figures and presented them in a variety of orientations, subsequently asking participants to duplicate the figures either via hand or a specially formulated keyboard. Those who handwrote the figures remembered them better and could reproduce them better, after a period of several weeks. Longcamp found, via FMRI recordings, that the state of the participants during learning trials showed activation in pathways long thought to be associated with geographic shapes; additional activations occurred in left Broca‘s and bilateral parietal lobes, areas related to handwriting, letter identification, and observation of actions. This bore out Longcamp‘s hypothesis that the movements a child uses when learning to write are specific precursors to the visual recognition and remembrance of letters and shapes, and that those selfsame sensorimotor activations would recur whenever the child tries to recreate the characters. Interestingly, it is then asked what this predicts for the future, in which school children are handed computers earlier and earlier.
Longcamp goes on to give the following illustration: pre-reader/new-reader child has trouble distinguishing between ‗b‘ and ‗d‘. There are no striking visual differences. However, the motor program called up to facilitate writing ‗b‘ is vastly different than writing ‗d‘; the difference is to such an extent that confusion seems difficult to fathom, yet it has long occurred in the general population, even prior to the elementary school introduction of computers. Synaptic motor connections are strengthened with time, and while such motor-visual interaction is not the source of all reading errors, it is clearly a significant source. What will increasing reliance on computers mean for the future of pedagogy?
As a matter of structure and comparison, Longcamp included an overtrained letters trial to ascertain the areas of brain activation that processed the novel characters. For the novel-handwritten /overtrained-letter trial: activation of the bilateral AIP, the left dorsal premotor, and the left postcentral regions, in addition to the total left-sided lateralization, were interpreted to clearly signify an interplay between visual processing and reactivated motor knowledge.
According to Longcamp, the most important finding of this study was that in Broca‘s left, stronger FMRI activation occurred for the handwritten than the type-written condition. He states that heretofore Broca‘s was thought of as a site for primarily speech production, but this study implies clear motor prominence as well.

Cognitive Longevity

Another important work in the discussion of brain/writing interplay is Danner and Snowdon‘s 2001 metanalysis of emotional content in a population of 180 handwritten nun autobiographies. Their work presupposed an association between early-life written emotional expression and late-life cognitive longevity. The dataset, consisting of autobiographies composed by prospective sisters from the order of Notre Dame in the late 1930s and early 1940s, is clearly outlined elsewhere, as is the structure of Snowdon‘s now classic longitudinal design.
Researchers examined existing data, previously rated for idea density (Snowdon, 1999) and fleshed out an earlier supposition (Snowdon, 1996) that the idea density rating scale could be correlated to emotional content within the autobiographies. If idea density equalled longevity, and emotional expression equalled idea density, then emotional expression now equalled longevity.
Before continuing, one should note that the majority of Snowdon‘s career has focused on the course of Alzheimer‘s Disease (AD) and prevention. AD is a neurodegenerative disease that affects millions of people. Currently, within the medical community, there is no accepted standard of what constitutes a neuropathologic diagnosis of AD, although strides are being made. It is still possible to look at an AD brain and a non-demented brain side by side and not quantifiably know the difference. As the symptoms of the disease are so insidious, this is difficult to fathom, yet it is the case.
Why? Upon autopsy, neurofibrallary tangles (NFTs), the primary indicator of AD, have been found both in individuals known to have manifested AD in life, and in those who never have. More important than the mere presence of NFTs appears to be the brain region in which they are located. According to Braak, NFTs rooted in the neocortex most assuredly lead to full-fledged AD manifestation. Additionally, Braak states that the first NFTs (which progressively grow, spreading throughout the brain) initially present in an individual‘s third decade.
The neocortex is the area of the mammalian brain key to expression of ideas. One may possess certain ideas on an intellectual basis (i.e. the world is a positive place overall) but unless one acts upon those ideas, ‗exercising‘ those parts of the brain, the protective benefit of idea expression is not experienced by the individual. So the nuns who wrote well in their 20s had likely been raised to do so, and it is likely they continued to do so throughout their lives, experiencing the protective benefits of said action (Danner & Snowdon, 2001; Mortimer, 2005).
In the current paper, Danner extrapolates from the above studies and other literature that if a negative event causes adverse effects, and one consciously chooses to reply from the perspective that the future holds only positivity, this generates an internal state that mutes any negative health-wise possibilities. For example, Frederickson spoke at length about cardiovascular system negative/positive response via activation from the autonomic nervous system, which is tied into individual emotional expression (1998). Danner recounts various other studies that support this idea.
But why EW? Tavakoli & colleagues (2009) and Smyth (1998) have both written to the validity of EW as a diagnostic measurement in that it allows and maintains a veil of privacy for the actor. Even when it is known what the end result of EW will be, for ex. evaluation by coders, the act of EW, the process that goes into creation and composition and completion, allows the actor to let go of an inherently more personal piece of emotional expression than would have otherwise been spoken aloud.
The Danner autobiographies were analyzed by three different coders and multiple statistical analysis were done, resulting in a clear positive correlation between early life positive emotional expression and late life longevity. This finding ran counter to certain other studies that found no protective benefit in positive affect. It stands to reason, however, that a continuing issue is truthfulness of self-report, as well as methods of ‗exercising‘ said self-expressive tendencies.


What, if anything, can EW offer clinicians? It is supposed here that, within the occurrence of EW, an actor recreates a TE, instilling or instituting a narrative framework of his own design. That framework could take the form of the traditional: a beginning, middle, and end, relative to me. The latter part of the sentence is key, as one might imagine. Any memory, relative to me, is more easily processed, stored and retrieved. Certainly this is what is at work in Longcamp‘s previously described novel figure recollection/learning task: a superior processing that is the end result of pairing a visual input with a motor action, all in relation to some pre-existing, familiar memory schema.
The EW narrative framework chosen could also be non-traditional: in re-experiencing a TE, there is no emotive context that is adequate in terms of healing. I cannott make this fit in the context of my life [because I am unwilling.] This seems more in tune with intense affective disorders, in that there appears to be a disregard for reality. The event triggers are not necessarily stronger, when compared with anything else, but with previous personality traits taken into consideration, addictive tendencies are likely to reveal themselves in the aftermath of TE and subsequent therapeutic recovery attempts. This is also in line with PTSD.
Without a doubt, the above has dribbled outside the lines of EW and into more conventional psychotherapy. The same question persists: how does the mind cope with stress, with TE? Psychotherapy and EW are different sides of the same coin. The ―context‖ aspect implies that play-acting might be as valuable, if not more so, a therapeutic method than EW. Again, there is a personality issue. Play-acting is performance oriented and thus is more intertwined with issues of self-consciousness and self-perception than is EW. According to Smythe and Tavakoli, EW is preferable to traditional psychotherapy in TE treatment because the emotive expression is veiled so as to shield the writer from vulnerability or embarrassment.
The narrative framework chosen in TE recovery may be traditional (story timeline), non-traditional (addiction as a ―plot‖ advancer), or time-unbound (as one might characterize recovering amnesiacs or potentially dissociative personalities). The point is that recovery is contingent on pre-existing personality traits (or organic conditions, in rare instances). The therapeutic relationship is there to guide the patient not only to the correct form of reinstitution (of integrating TE into traditional, healthy memory contexts, relative to me) but also, first, to recognize those pre-existing and harmful conditions that might derail successful TE narrative reconstruction.
In that vein, a pioneer in the field of applied psychology, Mihalyi Csikszentmihalyi compiled his theory of FLOW into a single text, in 1976. The book stands out as a guideline for life-wise application (taking into context TE, PE, and everything in between) regarding the entire scope of EW component traits and after-effects. Readers of the book were not told to journal on a regular basis, but the book itself is an exercise in traditional narrative construction—even if pen is never put to paper. How might one construct their life so as to best approach happiness? Do you know what your happiness is? From the ground up, Csikszentmihalyi builds a solid foundation for emotive reconstitution.
Interestingly, the issue of time raises its head in fairly early in FLOW. Csikszentmihalyi invented the Experience Sampling Method (ESM) in which he gave thousands of study participants a pager, and charged them with stopping whatever they were doing and writing in a journal whenever the pager randomly went off over a period of weeks. Results from this lead Csikszentmihalyi to claim that when people are most engaged in an activity, most interested in what they are doing—most involved in something that tests their skill level and challenges them—they lose track of time, and ―step outside themselves,‖ in a sense. They lose a recognizable perception of personal agency, at least until the challenge of the task itself has abated.
This may be where Klein and Boals‘ earlier question fits in: do TE and PE utilize separate brain pathways? As was said, any memory in relation to myself is more readily processed, more easily retrievable. People typically do not possess memory contexts or schemas for TE. There is no anticipation that bad things will occur, no hope—on the part of most people—that they will play the victim at some point in their lives. Having nothing in common with the ―Self,‖ TE is comprised of almost alien characteristics, in comparative terms. Multiple sessions of same-event EW corrects that by allowing the actor to construct a framework, a schema, a scenario in which TE is painted is ‗making sense‘ on some level, in relation to the actor. This is not lying to oneself. Rather, it is acknowledging the importance of the TE and creating adequate space for it.
It is important to note here that many TE victims report a slowing of time just during the course of the event, some report reliving episodes after the fact. This is not the lost sense of time that Csikszentmihalyi recounts, but rather a heightened awareness of time. This may, in fact, be another instantiation of ‗time-unbound‘ contexts, as related above.
That being said, does PE exert any kind of strong influence on the self? Based on the works referenced earlier, Klein supposed that PE made less of a cognitive impact than TE. Is time the reason? Csikszentmihalyi implies that PE bears no strong relation to the self. Not the event, PE. In the event, personal agency is, effectively, persona non grata.
Imagine Pollack straddling one of his canvases. In all likelihood, he would be unable to give a coherent account of his thought processes while under the influence of his art. He could certainly describe it, but the words would have little connection to reality for anyone other than Pollack. (Imagine the football player, breathless after a touchdown, describing his feeling of being ―in. the. zone.‖ These words are discernible to outsiders, but certainly not knowable). It is through seemingly erroneous verbal expressions that Pollack or footballer or whomever would access their internal schemas of who they are. This term ―painter‖ is a memory cue, the key that opens the lock for Pollack‘s visual ―excesses‖ and his complete range of compositional abilities. So while it is true that memory itself appears not to be propositional in nature, it does seem as though the word is the lock that fits into the key. Pollack does not envision himself ―a painter,‖ however he does envision himself painting. But he describes himself to others as a painter. Perhaps the great irony here is that society has ramped up its trafficking in labels, in exchange for an increase of efficiency.
To reiterate: imagine that a small child incessantly chirps ‗mommy‘ at everything because he does not know the right words yet. Perhaps he does know and fully understand ‗mommy‘; perhaps he‘s attempting to, on some level, place her attributes on these unknown things. If not for that reason, it may be that the child is acting, repeating an action solely under the motivation of achieving the same ecstatic response as before. Words are ‗packages‘ that may be reordered into a more favorable light. It is proposed here that the job of therapeutic EW is to teach others to use the written word to alter dangerous emotions and, at the same time, to remake faulty (or unhelpful) perceptions of events.
So, Pollack may find his flow as a painter, but one seemingly cannot paint in order to heal the scars of a TE. There is no way to place the event in a helpful memory context, no ―chronology‖ inherent to painting. Then again, memory itself is not chronological. Memory appears to be a visual entity, accessed by verbal cues, organized on seemingly hierarchic levels of decreasing importance to the self.
For that reason, one would still expect that, as people are such highly visual creatures, painting art therapy would be the better option than EW. There appears, however, to be a sensory neglect that comes on with time. One‘s appreciation of things visual fades from prominence as hours, even minutes pass. This degradation appears to occur on both a physical level and a psychological level. The oft-used illustrative example: can you name all the roads you cross on your way to work? No. Vision is the key to early learning, to memory contextualization, even to adult learning so that any time a person encounters something new, visual examination is often the weightiest consideration. This vision/motor-learning connection was referenced in Longcamp.
In conclusion, Klein asked whether EW could facilitate successful memory organization. What is disorganization? How should clinicians conceive of memories that exist outside of appropriate contexts? Seemingly, as dangerous as a ―standout memory‖ can be, are there not important inherent factors to be assessed? Is memory organization, then, similar to homeostasis, whereby no ripples occur whatsoever?
Considerations for future studies should include concerns for EW‘s effectiveness in treating major affective disorders. Additionally, what is the potential relationship between self-perception and, for instance, major depressive disorder? For that matter, what of self-perception and major anxiety disorder? Will self-image disorder one day be viewed as a component of major affective disorders? More simplistically, what about self-image/ self-perception problems in amnesiacs?

Belluck, Pam. ―Nuns offer clues to Alzheimer's and aging.‖ The New York Times, 07 May 2001, sec. A, p.1.
Danner, D., Snowdon, D., & Friesen, W. (2001). Positive emotions in early life and longevity: Findings from the nun study. Journal of Personality and Social Psychology, 80(5), 804-813.
Foa, E., & Kozak, M. (1986). Emotional processing of fear: Exposure to corrective information. Psychological Bulletin, 99, 20-35.
Foa, E., Steketee, G., & Rothbaum, B. (1989). Behavioral/cognitive conceptualizations of post-traumatic stress disorder. Behavior Therapy, 20, 155-176.
Fredrickson, B. L., & Levenson, R. W. (1998). Positive emotions speed recovery from the cardiovascular sequelae of negative emotions. Cognition and Emotion, 12, 191-220.
Kelley, J. E., Lumley, M. A., & Leisen, J. C. C. (1997). Health effects of emotional disclosure in rheumatoid arthritis patients. Health Psychology,16, 331-340.
Lemonick, Michael D. and Park, Alice. The Nun Study Time, 14 May 2001, p.54.
Longcamp, M., Boucard, C., Gilhodes, J., Anton, J., Roth, M., Nazarian, B., et al. (2008). Learning through hand- or typewriting influences visual recognition of new graphic shapes: Behavioral and functional imaging evidence. Journal of Cognitive Neuroscience, 20(5), 802-815.
Lutgendorf, S. K., Antoni, M. H., Kumar, M., & Schneiderman, N. (1994). Changes in cognitive coping strategies predict EBV-antibody litre change following a stressor disclosure induction. Journal of Psychosomatic Research, 38, 63-78.
Manning, Anita. ―Nuns' story of aging.‖ USA Today, 14 May 2001, sec. D, p.6.
Mortimer, James A.; Borenstein, Amy R.; Gosche, Karen M. & Snowdon, David A. (2005). Very Early Detection of Alzheimer Neuropathology and the Role of Brain Reserve in Modifying Its Clinical Expression. The Journal of Geriatric Neurology and Psychology, 18: 218-227.
Pennebaker, J. (1989). Confession, inhibition, and disease. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 22, pp. 211-244). New York: Academic Press.
Pennebaker, J. (1993). Putting stress into words: Health, linguistic, and therapeutic implications. Behavioral Research Therapy, 31, 539-548.
Pennebaker, J. W., Mayne, T., & Francis, M. (1997). Linguistic predictors of adaptive bereavement. Journal of Personality and Social Psychology, 72, 863-871.
Smyth, Joshua M. (1998). Written Emotional Expression: Effect Sizes, Outcome Types, and Moderating Variables. Journal of Consulting and Clinical Psychology, 66(1): 174-184.
Snowdon D.A., Greiner L.H., Mortimer J.A., et al. (1997). Brain infarction and the clinical expression of Alzheimer disease: The Nun Study. JAMA 277:813–7.
Tavakoli, Shedeh; Lumley, Mark A.; Hijazi, Alaa M.; Slavin-Spenny, Olga M. & Parris, George P. (2009). Effects of Assertiveness Training and Expressive Writing on Acculturative Stress in International Students: A Randomized Trial. Journal of Counseling Psychology, 56 (4): 590–596.
van der Hart, O., Steele, K., Boon, S., & Brown, P. (1993). The treatment of traumtic memories: Synthesis, realization, and integration. Dissociation,6, 162-180.
Weiss, Rick. ―Aging: New answers to old questions.‖ National Geographic, 192(5), November 1997, p.2-31.
Yogo, M., & Fujihara, S. (2008). Working memory capacity can be improved by expressive writing: A randomized experiment in a Japanese sample. British Journal of Health Psychology, 13(1), 77-80.