The Relational Prognosis: What a Conversation Analysis Measures (and Doesn't)

A prognosis isn't a prophecy

The question comes up often, mixed with hope and worry: "Can the analysis tell whether we'll stay together?" The honest answer is nuanced. A conversation analysis can surface trends — signals research (Gottman and others) has shown to correlate with stability or separation. But a trend isn't a destiny. A relational prognosis illuminates a probability, based on what is written today, not a future carved in stone.

Telling it from a prophecy is essential: that's what makes it a tool of clarity, not a source of anxiety or fatalism.

What the written word allows — and what it ignores

The written word offers precious material: a large, dated corpus of messages that reveals regularities (conflict management, positive/negative ratio, reciprocity, evolution). These regularities have real indicative value. But the written word doesn't capture everything:

• It sees only the written. A large part of the relationship plays out outside messages: in gestures, shared daily life, the nonverbal, lived experience off-screen.
• It doesn't know the context. A cold period may reflect a grief, an illness, external stress, not the bond itself.
• It doesn't measure the will to change. Yet that's often what decides the future, and it isn't in past data.
• It photographs, it doesn't predict. Trends describe the current state; they don't account for what a couple may decide to transform.

An honest prognosis therefore states probabilities with these limits attached — never a verdict.

What a prognosis honestly measures

• Correlated signals: presence of Gottman's horsemen, positive/negative ratio, repair capacity, balance of investment. These signals are, statistically, linked to couples' futures.
• Evolution trends: is the bond strengthening or degrading over time?
• Fragility zones: which dimensions threaten stability?
• Levers: what, if modified, would improve the prognosis.

The right reading isn't "we have a 60% chance of staying together" (an illusion of precision), but "here are the favorable signals, here are the at-risk ones, here's what to act on."

Reading a prognosis without submitting to it

• Take it as a signal, not a sentence. An unfavorable prognosis points to areas to work on, not a condemnation.
• Set it against your lived experience. The analysis ignores the off-screen; you don't. Confront the two.
• Focus on the levers. What matters isn't the displayed probability, but what it invites you to change.
• Beware the two excesses: neither fatalism ("it's hopeless") nor denial ("it means nothing").

It's in this spirit that the analysis from ScanMyLove reads your exchange history: not to render a verdict, but to illuminate trends and levers — a foothold for deciding, never a decision in your place.

From trend to decision

A well-understood prognosis serves action:

• Identify the at-risk signals and link them to precise, modifiable behaviors.
• Capitalize on the favorable signals: what works deserves to be cultivated.
• Decide knowingly. The data inform; the decision belongs to you and your partner.
• Get support. A psychological test sheds light on your part in the dynamic; and support at the practice helps turn a prognosis into a project, rather than endure it.

The written word illuminates, it doesn't decide

A relational prognosis is neither a crystal ball nor a sentence. It's an honest reading of trends, drawn from what is written, attached to its limits — it ignores the off-screen, the context, and above all your freedom to change. Where you dream of a definitive answer ("will we last?"), the analysis offers better: clarity about the present state and levers to act. And a couple that decides to act on its at-risk signals says more about its future than any percentage.

Gildas Garrec, CBT Psychotherapist in Nantes