A Legal & Social Analysis
Abstract
Uncertainty is not an anomaly in a criminal trial. It is one of its raw materials.
Witnesses remember incompletely. Experts work from limited inputs. Documents establish words without necessarily establishing intention. Psychological findings may be compatible with several historical pathways. Police officers observe facts and draw inferences from them. Courts must eventually transform all of this into findings of fact.
The difficult question for defence counsel is therefore not whether uncertainty exists. It almost always does. The question is how uncertainty is litigated.
This essay develops a comparative advocacy method inspired by the architecture of American cross-examination but applied within the procedural limits of a Spanish criminal trial. Its central proposition is that the defence does not necessarily need to prove that the prosecution’s interpretation is false. It may be enough to make visible the inferential distance between what the evidence actually establishes and what the court is being asked to conclude.
The task is not to manufacture doubt. It is to prevent inference from becoming invisible.
Keywords: criminal defence; cross-examination; reasonable doubt; expert evidence; inference; uncertainty; Spanish criminal procedure; trial advocacy; systems theory; evidential reasoning.
I. Uncertainty Is Not the Problem
Imagine the Seattle lawyer again.
He is sitting before a Spanish criminal court. The law is Spanish. The evidential standard is Spanish. The presumption of innocence is Spanish. The President controls the hearing under Spanish procedural law.
Nothing has been imported except a method of thinking about evidence.
The prosecution case appears coherent. There is testimony. There may be psychological evidence. There may be messages, medical documentation, subsequent conduct, police evidence or circumstantial material. No single element is obviously defective.
The defence faces a familiar temptation.
Find the mistake.
Find the contradiction.
Find the witness who is lying.
Find the expert who went too far.
Sometimes that evidence exists.
Sometimes it does not.
And sometimes the real defence lies somewhere else entirely.
It lies in the distance between what is known and what must be inferred.
That distance is uncertainty.
The forensic problem is not that uncertainty exists. The problem begins when uncertainty disappears from the courtroom language even though it remains embedded in the evidence.
A possibility becomes an explanation.
An explanation becomes the explanation.
A compatible finding becomes corroboration.
Corroboration becomes attribution.
And eventually the inferential chain is remembered only by its destination.
The trial lawyer’s task is to make the chain visible again.
II. The Courtroom Hates a Vacuum
Human beings are narrative organisms.
We do not naturally experience information as isolated propositions. We connect facts. We search for sequence, motive, cause and consequence.
This capacity is indispensable.
It is also dangerous.
A courtroom receives fragments:
A message.
A memory.
A symptom.
A delay.
A reaction.
A contradiction.
A medical finding.
A witness account.
An expert interpretation.
The mind begins to organise them almost immediately.
Once a plausible narrative emerges, subsequent information may be understood through that narrative. Ambiguous facts begin to acquire directional meaning because they fit the emerging whole.
This does not mean that the narrative is wrong.
It means that narrative coherence and evidential sufficiency are not identical concepts.
A story may be internally coherent and still depend upon propositions that were never independently established.
This is where criminal litigation becomes difficult.
The defence lawyer who simply offers a competing story may reinforce the idea that the court must choose between two complete narratives.
But sometimes the defence position is not:
“This other story is what happened.”
It is:
“The evidence does not allow us to travel that far.”
That is a different form of advocacy.
It is the litigation of uncertainty.
III. Begin With What Is Actually Known
The discipline begins with decomposition.
Take any important proposition in the case and ask four questions:
What was observed?
What was interpreted?
What was inferred?
What is ultimately being concluded?
The sequence can be expressed simply:
Observation → Interpretation → Inference → Conclusion
Those stages are related.
They are not interchangeable.
Suppose a psychologist observes anxiety during an evaluation.
That is an observation.
The psychologist interprets the pattern as compatible with trauma-related symptomatology.
That is an interpretation.
The symptomatology is associated with a particular event described by the examinee.
That is an inferential attribution.
The court is then invited, expressly or implicitly, to regard the psychological evidence as supporting the proposition that the historical event occurred as alleged.
That is a different conclusion again.
None of these transitions is necessarily illegitimate.
But each transition needs support.
The same structure appears outside expert evidence.
A witness sees a person leaving a building.
Observation.
The witness believes the person looked nervous.
Interpretation.
Nervousness is associated with consciousness of wrongdoing.
Inference.
The conduct is used as evidence that the person committed the offence.
Conclusion.
Again, the point is not that the conclusion must be false.
The point is that it did not emerge directly from the observation.
Something happened between them.
That “something” is where advocacy begins.
IV. Do Not Ask the Witness to Explain Your Case Away
Traditional examination sometimes gives away control unnecessarily.
Consider a psychological report containing the expression:
“Taking into account her degree of maturity…”
A conventional question may be:
“Doctor, could you explain what you mean by degree of maturity?”
The question sounds reasonable.
It may also be strategically disastrous.
The expert has been invited to explain.
She may speak for several minutes. She may place the expression within developmental psychology, clinical observation, professional experience and the totality of the assessment. She may clarify ambiguities that were more useful to the defence before she began speaking.
The Seattle lawyer approaches the same expression differently.
“When you refer to degree of maturity, did you measure that variable using a specific instrument?”
If yes:
“Which instrument?”
“What construct does that instrument measure?”
“What score was obtained?”
“Against what reference population?”
If no:
“So that expression reflects your clinical assessment rather than a separately quantified measurement?”
Then stop.
The objective is not to force the expert to withdraw the expression.
The objective is to establish its epistemic status.
Was it measured?
Was it observed?
Was it reported?
Was it interpreted?
Was it inferred?
Those are different categories of evidential information.
A courtroom should know which one it is dealing with.
V. Compatibility Is Not Causation
One of the most important disciplines in evidential reasoning is linguistic.
Certain expressions appear similar in ordinary speech but carry very different forensic consequences.
Compatible with ≠ indicative of ≠ corroborative of ≠ caused by ≠ proves
Consider the phrase:
“The symptoms are compatible with a traumatic experience.”
That may be scientifically unobjectionable.
But the phrase does not necessarily establish:
that trauma is the only possible explanation;
that the trauma was sexual;
that a particular historical event caused it;
that the event occurred exactly as later described;
or that a particular accused person was responsible.
Those propositions may eventually be supported by other evidence.
But they do not automatically reside inside the word compatible.
The problem arises when courtroom language migrates.
In the report:
“Compatible with.”
In oral argument:
“Consistent with.”
Later:
“Supports.”
Then:
“Corroborates.”
And eventually:
“Confirms.”
The evidence may not have changed.
The verb has.
And sometimes the verb carries the case further than the methodology ever did.
Litigating uncertainty therefore requires unusual attention to language.
A defence lawyer should listen not only for factual propositions but for changes in the level of certainty attached to those propositions.
VI. Litigate the Interfaces
Complex evidential cases often contain few obviously defective components.
The witness may be sincere.
The psychologist may be competent.
The medical report may be accurate.
The police officer may have described precisely what was observed.
The messages may be authentic.
The weakness may lie not inside those components but between them.
Consider the interfaces:
Statement → Assumed Historical Fact
Finding → Interpretation
Interpretation → Cause
Cause → Particular Event
Event → Legal Characterisation
Expert Opinion → Judicial Finding
Every arrow performs work.
The arrow should not be invisible.
This is especially important where several evidential systems interact.
A complainant gives an account.
The expert receives that account.
The expert identifies findings compatible with the account.
The findings are then presented as support for the credibility of the account.
That may or may not be methodologically legitimate depending on the particular evidence.
But the degree of independence matters.
If part of the interpretation derives from the same narrative that the interpretation is later said to support, the lawyer should identify the structure carefully.
Not to shout “circularity”.
Not to accuse the expert of bias.
Simply to ask:
“What supports this arrow?”
VII. Never Waste a Concession
There is a peculiar temptation in cross-examination.
Counsel finally obtains the answer he wanted.
And then destroys it by asking one question too many.
Suppose the exchange is:
“This instrument cannot determine whether a particular historical event occurred?”
“No.”
That proposition now exists in the evidential record.
The inexperienced advocate feels compelled to continue:
“So you accept that your conclusion does not prove my client committed the offence?”
The expert now has an invitation to repair everything.
She can explain that no single instrument is interpreted in isolation, that the evaluation was comprehensive, that professional judgment integrated multiple sources, that the court must consider the totality of the evidence and that the report was never intended to establish guilt independently.
Five minutes later, the clean concession has disappeared inside a perfectly respectable professional explanation.
The Seattle lawyer stops earlier.
The answer already exists.
He does not need the expert to deliver the defence closing argument.
He changes subject.
This discipline is fundamental to litigating uncertainty.
Uncertainty is rarely established by one spectacular admission.
It is accumulated through small, controlled propositions.
The closing argument will assemble them later.
VIII. Build the Record, Not the Speech
The most useful preparation question may be this:
“What propositions must exist in the evidential record when I sit down?”
That question changes preparation completely.
Instead of drafting fifty interesting questions, counsel identifies the propositions required for closing argument.
For example:
The psychological instrument measures symptoms, not historical authorship.
The relevant finding is compatible with more than one causal pathway.
The expert did not independently verify a particular factual assumption.
The expression “maturity” was clinically assessed but not separately quantified.
The witness cannot identify when a particular recollection first appeared.
The police officer observed conduct but cannot testify to the accused’s internal intention.
The message establishes the words used but not a unique interpretation of their meaning.
Once the propositions are defined, counsel works backwards.
What answer is needed?
What question is most likely to produce that answer?
What foundation must be established first?
What document fixes the witness if the answer changes?
Where should the examination stop?
This is where the American discovery mentality becomes useful even in a system without American discovery.
The mindset is reconstructive.
What did this witness know?
When did the witness know it?
From whom?
What was independently established?
What came from another participant?
What assumptions entered the analysis?
Which conclusions depend upon those assumptions?
The purpose is not to imitate American procedure.
It is to borrow American preparation discipline.
Months of preparation may produce twenty minutes of examination.
That is not inefficiency.
That is compression.
IX. The Defence Does Not Need a Complete Alternative Story
One of the most dangerous pressures on criminal defence is the perceived need to explain everything.
If the prosecution presents a complete narrative, the defence may feel compelled to provide another.
But an alternative historical reconstruction carries its own burden.
The defence may not know exactly why a witness behaved in a particular way.
It may not know what caused a particular psychological state.
It may not know why a message was sent at a particular moment.
It may not know precisely how an inconsistency emerged.
And it does not necessarily need to know.
If several explanations remain reasonably compatible with the evidence, the methodological point can sometimes stand without selecting one.
Suppose symptom Y could plausibly arise from pathways A, B, C or D.
The defence does not automatically need to prove B.
Its first task may simply be to establish that Y does not uniquely identify A.
This distinction matters enormously.
Otherwise defence counsel may transform a legitimate evidential limitation into a speculative alternative theory and then find himself defending a proposition he never needed to prove.
There is discipline in refusing unnecessary certainty.
X. Inferential Accumulation
Hard cases are rarely built upon one inference.
They are built upon layers.
Each layer may appear modest.
Each may appear individually reasonable.
The danger lies in accumulation.
Consider:
A witness gives an account.
The account is emotionally plausible.
Psychological findings are compatible with trauma.
Certain messages are interpreted as consistent with the account.
Subsequent conduct is regarded as understandable if the account is true.
A peripheral witness recalls something broadly compatible.
No individual proposition necessarily proves the historical allegation.
But together they create narrative density.
Narrative density can feel like evidential certainty.
The defence must therefore audit not only individual inferences but their interaction.
Does evidence B independently corroborate A?
Or does B derive part of its meaning from assuming A?
Does evidence C add new information?
Or does it merely become incriminating after being interpreted through A and B?
Does evidence D independently narrow the possibilities?
Or does it merely fit a narrative already constructed?
This is not an argument that circumstantial evidence is weak.
Circumstantial evidence can be extraordinarily powerful.
The question is whether the individual circumstances are genuinely independent and whether the inferential structure is being accurately represented.
Five pieces of evidence are not necessarily five independent confirmations.
Sometimes they are five expressions of the same informational source.
Sometimes they are genuinely independent.
The lawyer must know which.
XI. Cross-Examination as an Audit of Certainty
This suggests a different way of thinking about cross-examination.
Cross-examination is often described as an instrument for testing credibility.
It is also an instrument for testing certainty.
The lawyer can ask:
What is the source?
How direct is it?
Was it independently verified?
Does the methodology measure this proposition?
What alternative explanations remain possible?
Were they tested?
Does this conclusion depend upon information supplied by the person whose account it is later said to support?
Can the witness distinguish possibility from probability?
Can the expert distinguish compatibility from causation?
Can the police officer distinguish observed conduct from attributed intention?
Can anyone identify the precise evidential bridge?
This is not hostile examination.
It is an audit.
The strongest expert may actually make the best witness for this purpose.
A careful professional will often recognise methodological boundaries more readily than an advocate expects.
The defence does not need humiliation.
It needs precision.
XII. One Fact. One Question. One Boundary.
When the issue is uncertainty, compound questions are especially dangerous.
Consider:
“Doctor, since your tests cannot establish historical truth, the symptoms may have other causes, you relied partly on the complainant’s account and you did not exclude every alternative explanation, isn’t it true that your report cannot corroborate the accusation?”
The question contains methodology, causation, self-report, differential explanation, corroboration and the ultimate defence proposition.
The witness has been given an entire field in which to manoeuvre.
Better:
“The instrument measures symptoms?”
“Yes.”
“It does not identify the person who caused them?”
“No.”
“It does not determine when the underlying cause occurred?”
“No.”
“Your causal interpretation also incorporated information from the interview?”
“Yes.”
“That included the examinee’s account of the event?”
“Yes.”
Stop.
The defence has not proved the report wrong.
It has identified its perimeter.
That perimeter now belongs to the record.
XIII. The Architecture of Reasonable Doubt
Reasonable doubt is sometimes discussed as though it were a feeling produced at the end of trial.
For the advocate, it should be treated more structurally.
Doubt can have architecture.
It can arise because an essential proposition depends upon an unverified assumption.
Because a causal attribution exceeds what the methodology can establish.
Because several plausible pathways remain open.
Because supposedly independent corroboration derives from the same original source.
Because the interpretation of conduct is not the conduct itself.
Because a witness can establish what was seen but not what another person intended.
Because an expert can establish present psychological findings but not uniquely reconstruct a past historical event.
Because the prosecution narrative requires several inferential transitions and one of those transitions lacks sufficient support.
The defence advocate should therefore know, before closing argument, exactly where uncertainty resides.
Not vaguely.
Structurally.
Which proposition?
Which inferential step?
Which missing bridge?
Which alternative remains open?
Which degree of certainty has exceeded the evidential input?
Only then can reasonable doubt be argued as something more disciplined than rhetoric.
XIV. The Spanish Courtroom
The Seattle lawyer remains in Madrid.
He does not conduct an American deposition in open court.
He does not import Federal Rules of Evidence.
He does not transform the Spanish hearing into a jury trial.
The President controls the examination.
Questions must remain relevant.
Repetition may be stopped.
The court may intervene directly.
The procedural framework remains domestic.
But none of that prevents disciplined preparation.
Indeed, the Spanish format makes compression even more important.
The lawyer may have limited time.
The expert may already have delivered a detailed report.
The professional judges may have read the material in advance.
That makes theatrical confrontation less useful, not more.
Twenty precise propositions may be more valuable than an hour of argument with the witness.
The purpose of the examination is not to perform scepticism.
It is to place usable limits into the evidential record.
XV. When the Expert Is Right
The hardest case may be the one in which the expert is right.
The assessment was professionally conducted.
The instruments were appropriate.
The symptoms are genuine.
The interpretation is clinically reasonable.
The report contains no obvious methodological defect.
What then?
Then the defence asks a narrower question.
Not:
“Is the report correct?”
But:
“Correct for what proposition?”
An expert may be entirely correct in identifying a psychological condition and still be unable to determine uniquely which historical event caused it.
A physician may accurately diagnose an injury and still be unable to reconstruct every detail of its mechanism.
A police officer may accurately describe conduct and still be unable to identify subjective intention.
A witness may accurately remember one part of an event while remaining unable to establish another.
Accuracy within a domain does not remove the boundary of that domain.
This may be the most important lesson in difficult criminal litigation.
The defence does not always win by showing that evidence is bad.
Sometimes it must show that good evidence is being asked to do more than good evidence can do.
XVI. The Discipline of “We Do Not Know”
There is an uncomfortable sentence in criminal advocacy:
“We do not know.”
Lawyers resist it.
Experts resist it.
Witnesses resist it.
Courts understandably seek resolution.
But intellectual discipline sometimes requires precisely that answer.
We may know that a symptom exists without knowing its unique cause.
We may know that a message was sent without knowing its only possible meaning.
We may know that conduct occurred without knowing the internal state that produced it.
We may know that a witness remembers something without knowing every process through which that memory developed.
We may know that one hypothesis fits the evidence without knowing that no other hypothesis reasonably does.
“We do not know” is not an abdication of reasoning.
Used properly, it is the conclusion of reasoning.
And in criminal proceedings, the boundary between what is known and what remains unknown matters.
XVII. Litigating Uncertainty
The defence does not litigate uncertainty by creating confusion.
It does not need to suggest that everything is unknowable.
It does not need to attack science, psychology, memory, expert evidence or human testimony as inherently unreliable.
That would merely replace one form of overstatement with another.
The task is narrower.
Identify what the evidence establishes.
Identify what it does not.
Identify the inferential operations required to travel from one to the other.
Then test those operations.
Source.
Measurement.
Interpretation.
Alternative.
Attribution.
Independence.
Causation.
Scope.
Certainty.
And once the necessary propositions exist in the record, stop asking the witness to decide the case.
That is the court’s function.
The lawyer’s function is to ensure that the court can see where evidence ends and inference begins.
Sometimes the best cross-examination will not establish that the expert was wrong.
Sometimes it will establish something more modest.
And in a criminal courtroom, potentially more important.
The evidence may be sound. The inference may be plausible. But the distance between them still has to be proved.
That is how uncertainty is litigated.
