Companions
Challenge
Blog
←Back to Blog
Mental Health

Shadow Work, Honestly: What Jung Meant, What the Evidence Supports, and a 15-Minute Version That Won't Make Things Worse

ILTY TeamILTY Team•November 3, 2026•14 min read

In crisis? Call or text 988 in the US, 9-8-8 in Canada, or find your country's line at findahelpline.com.

It is a Tuesday lunch break and you are watching a pair of hands open a spiral-bound journal under a ring light. The prompt on the page reads something like: which trait do you judge most harshly in other people, and where do you have it? The comments say this one broke them. You bought the journal three weeks ago. It is on the nightstand, you have done four prompts, two of them left you feeling worse in a way you could not name, and you stopped.

That is shadow work as most people meet it: a Jung-flavoured prompt book, bought in a feed, worked through alone, late, with no stop time and nobody to tell. Shadow work contains one genuinely useful idea, which is Jung's, wrapped in a product nobody has tested; whether it helps or hurts you depends on what you point it at, how long you sit in it, and whether you do it alone.

What is shadow work in psychology?

Shadow work means noticing and owning the parts of yourself you have disowned, and the idea belongs to Carl Jung. In his 1937 Terry Lectures at Yale, published as Psychology and Religion, Jung wrote: "Everyone carries a shadow, and the less it is embodied in the individual's conscious life, the blacker and denser it is" (Collected Works 11, par. 131). Elsewhere he called the shadow "the thing a person has no wish to be" (CW 16, par. 470).

Two of Jung's own points get lost in the TikTok version. First, the shadow is not a monster. In the same volume he called it "merely somewhat inferior, primitive, unadapted, and awkward; not wholly bad" (CW 11, par. 134), and in Aion (1951) he wrote that it "does not consist only of morally reprehensible tendencies" (CW 9ii, par. 423). Your shadow can hold hidden ambition, or a talent that made someone at home uncomfortable.

Second, the mechanism, which is the useful part. In Aion, Jung called the shadow "a moral problem that challenges the whole ego-personality" (CW 9ii, par. 14) and described projection: the quality you refuse in yourself gets seen in someone else, so the cause of the emotion appears to lie in the other person (par. 16). That is why the heat of a reaction is the tell. The colleague whose self-promotion makes you furious: the size of the feeling, not the colleague, is where the information is.

What Jung did not write was prompts. His own method was active imagination, recorded in the notebooks published in 2020 as The Black Books 1913-1932, the source of about half of The Red Book according to the Philemon Foundation. "Shadow work" as a named practice with tools arrives with the popularisers: the 1991 anthology Meeting the Shadow, edited by Connie Zweig and Jeremiah Abrams, has a section titled "Tools for Shadow Work."

How shadow work became a TikTok product

The journal under the ring light is most likely The Shadow Work Journal by Keila Shaheen, a self-help author who self-published it and sold it through TikTok Shop. The numbers: The Atlantic reported in September 2023 that it had sold 290,000 copies through TikTok Shop since that April. By May 2024 The New York Times, citing the publisher, put sales at over a million copies (both figures as carried by Publishers Marketplace).

The self-published edition also had problems a traditional publisher would normally catch. Satya Doyle Byock, a licensed professional counselor at the Salome Institute of Jungian Studies, wrote in June 2024 that it ran a photograph of Carl Schurz labelled as Jung, credited Freud and Klein with developing Jung's theory, and used a "Jung" quote she could not find in the Collected Works. That is her account. The broader point: a product sold in a feed has the quality control of a product sold in a feed.

That does not make the prompts useless. It makes them Jung-flavoured rather than Jung, and untested.

What does the evidence actually support?

We found no trial of shadow work. A search of Europe PMC for "shadow work" and Jung returns four papers, none of them a randomized trial, a systematic review or an outcome study. The honest question is what we know about its ingredients.

Writing about hard things helps a little

The ancestor of every prompt journal, and of journaling for mental health generally, is a 1986 experiment by James Pennebaker and Sandra Beall. Four groups wrote about different things; only the group that wrote about both the facts and the feelings of a traumatic event made fewer illness-related doctor visits in the weeks that followed. Neither alone did it.

Twenty years and 146 randomized studies later, Frattaroli's 2006 meta-analysis put the average effect of this kind of disclosure at r = .075: real, and small. In clinical samples she called the results "quite mixed," and one 2002 study she reviewed found disclosure increased doctor visits in a small sample of men in PTSD treatment. A 2023 meta-analysis by Guo of 31 randomized trials (4,012 people) found expressive writing reduced depression, anxiety and stress by g = -0.12, with sessions spaced one to three days apart doing better than other spacings. That is the size of "works."

Naming the feeling helps, but not as advertised

The "backed by neuroscience" line in shadow work videos usually rests on studies like Matthew Lieberman and colleagues' 2007 experiment: 30 people in an fMRI scanner. Putting a word to an emotional face reduced amygdala activity and increased activity in the right ventrolateral prefrontal cortex. That is a real result about labelling faces on a screen, not a study of journaling, of Jung, or of your childhood.

Self-compassion is the part the prompts skip

Shadow work asks you to look at the worst thing about yourself; the next sentence matters more than the looking. Kristin Neff's 2003 model describes self-compassion as self-kindness, a sense of common humanity, and mindfulness. Ferrari and colleagues' 2019 meta-analysis of 27 randomized trials of self-compassion interventions found effects of g = 0.75 on self-compassion, 0.66 on depression, 0.57 on anxiety and 1.37 on rumination (the authors warn that publication bias may be present). Even discounted, it belongs at the end of every shadow session.

Where it goes wrong: rumination

The failure mode is a quiet loop. Susan Nolen-Hoeksema and colleagues' 2008 review found that rumination "enhances negative thinking, impairs problem solving, interferes with instrumental behavior, and erodes social support," and predicts the onset of depression. An open-ended prompt like "what is wrong with me?" with no stop time is a rumination invitation in a nice font. It is the same reason journaling backfires for some people and thinking harder about who you are tends to make the question murkier. The science of rumination covers how to interrupt the loop.

Three groups who should not do shadow work alone

Anyone with a trauma history, anyone in an active episode of depression or panic, and anyone whose earlier attempts at writing about themselves left them worse. The Cleveland Clinic (January 2026) advises stabilizing panic, insomnia, anxiety or depression first and warns that "you might feel worse before you feel better."

The trauma lesson comes from psychological debriefing: a single session, soon after the event, talking it through in detail. A 2002 Cochrane review (Rose, Bisson, Churchill and Wessely; 11 trials) found it neither reduced distress nor prevented PTSD, and in one trial the debriefed group had higher odds of PTSD a year later (odds ratio 2.88). The authors' conclusion: "Compulsory debriefing of victims of trauma should cease." Getting it all out, early and unstructured, can be worse than nothing.

Writing about trauma does work in one form. In a 2018 trial in JAMA Psychiatry, Denise Sloan and colleagues randomized 126 adults with PTSD to five sessions of Written Exposure Therapy or twelve of cognitive processing therapy; the writing was noninferior, and far fewer people dropped out (6.4 versus 39.7 percent). But it was clinician-delivered: a trained person set the structure and was there for what came up. NICE (NG116, 2018) recommends trauma-focused CBT or EMDR for PTSD, and the VA's National Center for PTSD lists CPT, EMDR and prolonged exposure as having the strongest support; none of them is a journal on a nightstand.

Red flags that mean stop: flashbacks, losing time or feeling unreal while writing, replaying the same scene for hours, lost sleep, or writing that turns hopeless. If it turns to not wanting to be here, call or text 988 in the US, 9-8-8 in Canada, or find your country's line at findahelpline.com.

How do you start shadow work? A 15-minute version

Point it at this week, not your childhood. Time-box it. End in self-compassion and one action.

  1. Pick a reaction, not a wound (2 minutes). One moment from the last seven days when your reaction outran the event: the hour of fury over a two-line message. Jung's point about projection (CW 9ii, par. 16) is that this heat is the signal.
  2. Facts first, then the feeling (5 minutes). What happened, in plain sentences, then what you felt, with a specific word. "Annoyed" is a category; "humiliated" is information. Pennebaker's original result needed both.
  3. Ask one shadow question (4 minutes). "Where do I do a version of this?" or "What would it cost me to admit I have some of this?" One question, not the page of them.
  4. Finish with self-compassion, on purpose (2 minutes). Neff's three parts: I can be decent to myself about it; this is a human thing; I can look at it without becoming it. In the trials, this is what moved rumination most.
  5. Convert it into one action before Thursday (2 minutes). One sentence, one date. Nolen-Hoeksema found rumination "interferes with instrumental behavior"; an action is the exit.
ElementShadow work as soldA sane version
TopicChildhood wounds, earliest memoriesThis week's oversized reaction
LengthOpen-ended, until something comes up15 minutes on a timer, stop mid-sentence
FrequencyDaily, in a streakEvery one to three days, the spacing that did best in Guo's 2023 meta-analysis
EndingInsightSelf-compassion, then one dated action
CompanyAlone, lateOne line said to a witness

Two guardrails sit outside the clock. Read-back, the same test this site uses for journaling: half an hour after you stop, are you clearer or heavier? Heavier twice in a row means this is not your tool right now; journaling that makes you feel worse is a known pattern with alternatives. Witness: say one sentence of what you found to someone who will not flinch. A page nods along; a person can ask whether the story is true, which is why checking your story against the evidence works better with company.

What are examples of shadow work prompts that won't make things worse?

Each is tied to a mechanism above, and none asks about trauma. One per session.

  1. Who irritated me most this week, and what exactly did they do? Projection (CW 9ii, par. 16): start from the heat, then get specific.
  2. Which trait would I be most offended to be accused of, and when did I last do a small version of it? Jung's "thing a person has no wish to be" (CW 16, par. 470), at a size you can hold.
  3. Which compliment do I always deflect, and what would accepting it commit me to? The shadow is "not wholly bad" (CW 11, par. 134); disowned strengths count.
  4. What did I say I was too busy for this month, and what was the real reason? Self-deception hides in true sentences; the second half is the shadow.
  5. What would I say to a friend who admitted this about themselves? Self-compassion by proxy (Neff, 2003); it stops a session becoming a verdict.
  6. What is one thing I will do differently before Thursday because of what I wrote? The exit from rumination into action (Nolen-Hoeksema, 2008).

Skip, unless a clinician is in the room: "describe in detail" prompts that end in a memory you have never told anyone, anything about what your parents failed to give you, anything asking for your "deepest wound." Those are exposure exercises; exposure is a treatment with a protocol, not a journal page.

Frequently asked questions

Is shadow work dangerous? Kept to recent reactions and 15 minutes, for most people no; the realistic risk is a rumination loop. It is riskier with a trauma history, active depression or panic, where unstructured digging can open things nobody is there to help close. Flashbacks, dissociation or hopelessness while writing mean stop and talk to a professional. If it turns to not wanting to be here, call or text 988 in the US, 9-8-8 in Canada, or find your country's line at findahelpline.com.

Is shadow work the same as therapy? No. Shadow work is an untested self-help practice; therapy is delivered by a trained person who sets the structure and is responsible for what comes up. Jungian analysts do work with the shadow, but inside a therapeutic relationship, which is not a prompt book.

How do I find my shadow self? Watch for reactions bigger than their cause, compliments you cannot accept, and traits you would be most insulted to be accused of. Jung's view was that the disowned quality shows up projected onto other people, so the heat of the feeling is the map.

What is a shadow work journal? A prompt book organized around Jung's idea of the shadow. The best known, Keila Shaheen's The Shadow Work Journal, sold 290,000 copies through TikTok Shop by September 2023 (The Atlantic) and over a million by May 2024 (publisher's figure). None we could find has been tested in a trial.

Sources and further reading

ILTY is built for steps three to five. Tell it the reaction; the companion asks about it instead of handing you a page of prompts. Mr. Relentless will ask where you do a version of the thing you hate, and because ILTY remembers past conversations, he will ask next time whether you did what you said. Every conversation ends with two to four action items. It is not therapy, it does not treat trauma, and it is not for crisis. It is a witness that talks back at 11:40pm. Start the 7-day trial.

#shadow work#shadow work prompts#jung#journaling#self-awareness#rumination

Share this article

ILTY Team

ILTY Team

AI Mental Health Companion

Building an AI companion that actually helps with your mental health.

Get mental health insights in your inbox

No fluff, no toxic positivity — just what actually helps.