Joseph Murphy on forgiveness, compared with the research (REACH model)

What Murphy wrote about forgiveness

Murphy’s chapter rests on a theological claim before a psychological one: God, or Life, does not punish. Guilt and resentment, in his view, are self-inflicted, and so they can be released. His summary offers a play on words: “To forgive is to give something for.” What you give, he explains, is goodwill: “love, peace, joy, wisdom, and all the blessings of life”, wished for the other person instead of the grudge.

His test of whether you have forgiven, which he calls “the acid test”, is practical. You wish the other person well, and you keep doing so until, when the memory comes back, there is “no sting left in your mind”. The chapter also contains a named “Technique of Forgiveness”: quiet the mind, relax, think of God’s love for you, then affirm that you fully and freely forgive the person and release them. Once you have forgiven, he adds, there is no need to repeat this prayer: whenever the person comes to mind, you simply wish them peace.

Two things in this deserve respect. Murphy does not make forgiveness depend on the other person’s apology, and he treats it as a process you return to, not a single moment of willpower. Both ideas are close to how researchers describe forgiveness today.

What forgiveness is, and what it is not

The Greater Good Science Center at Berkeley summarises the research definition as “a conscious, deliberate decision to release feelings of resentment or vengeance toward a person or group who has harmed you, regardless of whether they actually deserve your forgiveness.”

The same page is clear about what it does not mean: forgiving does not mean forgetting, condoning or excusing; you do not gloss over or deny how serious the harm was; and it does not oblige you to reconcile with the person.

Murphy says something close himself: forgiving someone, he writes, does not necessarily mean liking them or wanting to spend time with them. The research adds a caution. James McNulty followed 72 newlywed couples over their first two years of marriage (2008): more forgiving spouses stayed more satisfied when their partner was rarely negative, but their satisfaction fell faster when the partner was often negative towards them. The study is correlational and small, so read it as a caution rather than a rule. It fits the definition above: forgiving someone who keeps hurting you does not oblige you to stay close to them.

Worthington’s REACH model, step by step

Everett Worthington, a psychologist and forgiveness researcher (he co-wrote the 2014 meta-analysis below), built a five-step model called REACH. His own site describes it as follows.

• R, Recall the hurt: face the fact that you were hurt, without treating yourself as a victim or the other person as a villain.
• E, Empathize: try to see the event from the other person’s side, for instance by imagining them in an empty chair, then sitting in their chair.
• A, Altruistic gift: offer forgiveness as a gift, remembering times you were forgiven yourself.
• C, Commit: make it concrete, for instance with a note to yourself such as “Today, I forgave (name) for hurting me.”
• H, Hold onto forgiveness: when doubts or the old feelings come back, reread what you wrote.

What the meta-analysis found

In 2014, Nathaniel Wade, William Hoyt, Julia Kidwell and Everett Worthington ran a random-effects meta-analysis of interventions designed to promote forgiveness, each led by a professional: 54 published and unpublished reports, with 53 effect sizes after the intervention (2,323 participants) and 41 at follow-up (1,716 participants).

Compared with no intervention, people who went through a forgiveness programme reported more forgiveness, with a standardised mean difference of 0.56 (95% confidence interval 0.43 to 0.68). Compared with other interventions, the advantage was 0.45 (0.21 to 0.69). They also reported more hope. Longer programmes, and those delivered to individuals rather than groups, went with larger effects; once these and the severity of the offence were taken into account, the differences between models faded. One precision for REACH: no study in the analysis tested it one-to-one, so the one-to-one advantage was shown most clearly for Robert Enright’s process model.

Read the limits too. Every programme was led by a professional (it was a condition for inclusion), longer ones did better, and forgiveness was self-reported. A few minutes in an app at bedtime is a much lighter dose, so the meta-analysis tells you the direction of travel, not what one night will do.

Murphy and the research, side by side

Put next to REACH, Murphy’s chapter covers some steps well and skips others.

A gentle nightly version, in the app

Do not picture the other person apologising: their choices are theirs. Picture yourself at ease with the memory. For example: “I walk past the café where we argued, and my shoulders stay loose; I notice the smell of coffee, not the old words.”

In Dream Architect (options counted in the app on 20 October 2026), choose an area such as Inner life or Calm in “Where it shows” (full flow), write that scene under “Write my own moment” in “The moment”, and in “The feeling” pick Relief or Ease, the two that fit this kind of scene best among the 5. At “Sleep on it”, tap “I close my eyes”: your moment stays on the dark screen, then fades into a slow breathing ring.

The next morning, the morning check-in asks: “Last night you relived this moment. This morning, did anything remind you of it?” Its last step, “One line”, is optional, and it is where REACH’s written commitment fits: one sentence such as “I wish them well, from a distance.” Your morning lines stay in the Journal, so step H, rereading, is one tap away.

If the memory brings up more than you can hold at bedtime, stop the practice there and talk to someone you trust or a qualified counsellor. Forgiveness for a serious harm is often a long road, and the research programmes were built for it.

Questions people ask next