A person gives an account of events. Some time later, they give another statement and certain details have changed. In legal proceedings, those differences matter. But the existence of an inconsistency and the explanation for that inconsistency are two different questions.

Information may appear that was not mentioned before. An element may be placed at a different point in time. A sequence may be described differently. There may even be contradictions between versions.

Immediately interpreting a difference as evidence of deception can be as premature as ignoring it. Research on repeated eyewitness interviews shows why consistency should be treated with greater precision: details newly reported in a later interview can remain highly accurate, and consistency does not operate as a simple indicator of truth or deception.

Memory does not work like a recording

When we remember an event, we do not replay an exact and unchanging copy. Remembering is reconstructive. What can eventually be communicated depends on how the event was perceived, the attention available at the time, the passage of time and the conditions under which the information is later retrieved.

What happens afterwards also matters. Decades of research on the misinformation effect have shown that later exposure to misleading information can affect memory for an event. This means we may need to consider not only what a person remembers, but also what information they may have received between the event and a later statement.

The fact that two accounts are not identical does not, by itself, tell us why they differ.

Not all inconsistencies mean the same thing

Describing a statement as “inconsistent” as though it were a single category oversimplifies the issue. The literature distinguishes different forms of consistency: within the same account, across successive accounts, between different people, or between a statement and the available evidence.

For legal strategy, the concrete relevance of what changes also matters. A difference concerning a peripheral detail and a contradiction concerning a fact with direct evidential significance may require very different strategic responses. That does not mean either one automatically reveals deception; it means they should not be analyzed as though they were equivalent.

How the statement was obtained also matters

The content of an answer cannot be completely separated from the context in which it was obtained. A free narrative does not necessarily provide the same information as a response to a closed question, and a neutral question is not equivalent to one that introduces information or directs attention toward a particular alternative.

For that reason, when two statements differ, reviewing only the answers may leave out a relevant part of the problem. It may be necessary to examine what was asked, how it was asked, what information was available to the person at the time and what occurred between one statement and the next.

A statement does not exist in isolation from the evidence

An inconsistency becomes more informative when it can be tested. Documents, communications, records, images, other people’s statements or other independent evidence may confirm certain points, challenge them or open alternative explanations.

The objective should not be to decide intuitively whether someone “seems credible”, but to distinguish what we observe from the possible explanations for what we observe.

A difference may relate to memory processes, the way an earlier statement was obtained, information incorporated later, a different interpretation of the question, deliberate omission or deception. Some hypotheses may be ruled out when tested against the evidence. Others may not.

An inconsistency is a data point. Its explanation is a hypothesis.

What does this change for legal strategy?

An inconsistency can do more than challenge a statement. It can also show where to look more closely.

It may identify what needs to be tested against independent evidence, which documents should be reviewed, which part of an earlier account requires context, or where a line of interviewing or questioning may be particularly relevant.

At that point, a contradiction stops being only a credibility issue and becomes information capable of guiding the analysis.

Observe before interpreting.

Separating the data from its explanation reduces the risk of turning a hypothesis into a conclusion too early.