Death for Stealing a Hat? Three Years for Assaulting a Child? What Does This Have to Do with Interoperability?

By: John F Groom

In London in 1826, a person could be sentenced to death for stealing a hat, while another person might receive three years in prison for attempting to rape a child. How could a legal system conclude that stealing a piece of clothing deserved death while a violent sexual assault deserved only a few years of imprisonment? The answer is not simply that people in the nineteenth century had different moral values. Nor is it that they considered property more important than children, although property rights certainly enjoyed extraordinary legal protection. A more fundamental explanation is that the criminal justice system was not interoperable.

That may seem like an odd word to apply to English criminal law two centuries ago. We generally associate interoperability with computers, databases, software, and artificial intelligence. But the underlying concept is much broader. Interoperability is the ability of different components of a system to exchange information, understand one another, and work together toward a common purpose. The English criminal justice system of 1826 had hundreds of laws. What it lacked was a consistent way to compare them.

The Bloody Code

By the early nineteenth century, England had accumulated more than 200 capital offenses under what historians call the Bloody Code. Many of these offenses involved property. Certain forms of theft, burglary, robbery, forgery, and damage to property could result in a death sentence. The circumstances of the offense, the value of the property, and the particular statute under which a defendant was prosecuted could make enormous differences.

Rape was also a capital offense. Attempted rape, however, was treated differently and could result in a relatively short prison sentence. The result was a system in which the legal consequences of different crimes often bore little apparent relationship to their relative seriousness.

There is an important qualification. A death sentence did not necessarily mean execution. Pardons, commutations, and transportation frequently intervened. Juries sometimes deliberately undervalued stolen property to avoid triggering capital punishment. But that only reveals another problem: when the formal rules produced outcomes that participants considered unreasonable, judges, juries, and officials had to find ways around them. The system required human beings to compensate for its structural inconsistencies.

How Did This Happen?

The criminal law had not been designed from the beginning as one comprehensive, internally consistent framework. Instead, it developed incrementally over centuries. A particular problem arose, Parliament passed a law, another problem arose, and Parliament passed another law. Existing laws were amended, judicial precedents accumulated, and new offenses were created. Each addition might have seemed reasonable within its own narrow context.

A law protecting valuable commercial property might have been justified as necessary to discourage theft. A separate law governing sexual offenses might have reflected centuries-old legal definitions distinguishing completed crimes from attempts. But who was responsible for comparing the two? Who examined the entire criminal code and asked whether the punishment for stealing a hat was proportionate to the punishment for attempting to rape a child?

There were reformers who raised precisely such questions, including in parliamentary debates during 1826. But the legal structure itself had no reliable mechanism for continuously identifying and correcting these inconsistencies. The laws existed together without necessarily functioning as parts of a coherent whole. They were connected by jurisdiction and enforcement, but not by a common evaluative architecture.

The Difference Between Connection and Interoperability

Imagine a modern organization with 200 departments. Each department has its own database, terminology, operating procedures, performance measures, and rules for making decisions. All 200 departments report to the same chief executive. They use the same accounting system and may even occupy the same building. Does that make them interoperable?

Not necessarily. Suppose one department classifies a customer as high-risk because of a particular financial characteristic, while another classifies the same customer as low-risk using entirely different criteria. A third department refuses to serve the customer, while a fourth offers preferential treatment. Each department might be following its own rules perfectly, yet the organization as a whole is producing contradictory results.

The problem is not necessarily that any individual rule is defective. The problem is that the organization lacks a shared framework for interpreting information, comparing decisions, and identifying inconsistencies. That is substantially what happened in English criminal law. A statute governing theft and a statute governing sexual assault were both part of the criminal justice system, but their punishments were not derived from a common, consistently applied assessment of harm, intention, coercion, vulnerability, and risk. The laws could coexist without being meaningfully comparable.

What Would an Interoperable System Look Like?

Consider an alternative in which every criminal offense is represented using a common set of characteristics. These might include the offender's intention, the harm actually caused, the harm reasonably risked, the use of force, the vulnerability of the victim, the degree of planning, and any aggravating or mitigating circumstances.

Theft, assault, fraud, rape, and murder would remain different offenses. They would not need identical legal definitions or identical punishments. But they would be described using enough shared information that comparisons became possible. The system could then ask whether a punishment assigned to one offense was consistent with punishments assigned to other offenses involving comparable or greater levels of harm. It could flag apparent contradictions, identify laws that had become inconsistent because one was amended while another remained unchanged, and show how the same conduct was treated differently under different statutes. Most importantly, it could make the consequences of changing one part of the system visible elsewhere.

That would not eliminate disagreement. Reasonable people can disagree about how much weight to give property rights, physical injury, psychological suffering, intention, deterrence, and rehabilitation. Interoperability cannot decide moral questions by itself, but it can expose contradictions that otherwise remain hidden inside separate rules and institutions. Interoperability does not guarantee good judgment. It makes consistent judgment possible.

The Problem Extends Far Beyond Criminal Law

Consider healthcare. One physician treats a patient's blood pressure, another treats the patient's sleep disorder, and a third treats a musculoskeletal problem. Each has access to different information, uses different measures of success, and recommends interventions based on a particular specialty. All three physicians may be highly competent, yet the patient can receive recommendations that conflict with one another or fail to account for the patient's overall health, priorities, and daily functioning. The problem is not necessarily medical knowledge. It is the absence of a sufficiently integrated view of the person.

Government regulation presents a similar challenge. A transportation agency establishes one set of requirements, an environmental agency establishes another, and a housing authority imposes a third. Each may be pursuing a legitimate public objective, but no one necessarily evaluates the combined effect of all three on the cost, feasibility, or quality of a proposed development. A project that satisfies each agency's individual requirements may nevertheless become economically impossible or produce a worse overall outcome.

Education has similar problems. Students are evaluated through examinations, grades, attendance, standardized tests, extracurricular activities, and teacher assessments. These measurements may be collected in separate systems that cannot adequately interpret one another. Businesses accumulate comparable inconsistencies in their hiring policies, compensation systems, performance measures, procurement procedures, and customer relationships. In every case, the underlying problem is similar: individual components make decisions without sufficient understanding of the larger system in which those decisions operate.

Why Artificial Intelligence Changes the Possibilities

For most of human history, creating a genuinely interoperable system was extraordinarily difficult. Information was recorded in different formats, using different terminology, for different purposes. Even when records were available, comparing thousands of rules and identifying their interactions required enormous amounts of human effort. Computers made it easier to store and retrieve information. Databases made it easier to organize information. Networks made it easier to exchange information. But none of these developments automatically solved the problem of meaning. Two systems might exchange data perfectly while interpreting the same information differently.

Artificial intelligence creates the possibility of going further. An AI system can examine information from different sources, identify the objects and concepts being described, interpret relationships, and compare decisions across domains that were never designed to work together. In principle, it could examine an entire criminal code and identify inconsistent punishments, conflicting definitions, redundant provisions, and unexpected consequences of proposed amendments. It could perform similar analyses across healthcare, education, business operations, and government regulation.

This requires more than simply feeding documents into a language model. Reliable interoperability also requires persistent identification of the relevant objects, meaningful descriptions of what those objects represent, records of their relationships, and mechanisms for testing whether the resulting interpretations are correct. But the important change is that we can begin to treat collections of independently developed rules, records, and processes as components of larger systems and evaluate those systems as wholes.

The Greater Opportunity: Discovering What We Have Never Compared

The most interesting benefit of interoperability may not be making existing systems more efficient. It may be discovering relationships that nobody previously thought to examine The English criminal code provides a simple example because the inconsistency is so striking. Death for certain property crimes and three years for an attempted sexual assault invite an immediate comparison. But most inconsistencies are less obvious. They may involve information recorded by different organizations, at different times, in different countries, and for entirely different purposes.

One dataset might describe physical exercise, another might describe medical outcomes, and another might record occupational activity, environmental conditions, or individual behavior. Viewed separately, each dataset has a particular purpose. Viewed together, they might reveal relationships that none of their original creators anticipated. The same is true of laws, patents, scientific research, business practices, and historical records. Interoperability allows information to acquire value beyond the purpose for which it was originally collected. It creates the possibility of asking questions that could not previously be formulated, much less answered.

From Accumulated Rules to Intelligent Systems

There is a broader lesson in the English criminal justice system of 1826. Many of the institutions we depend on were never designed as complete systems. They evolved. People encountered problems and created solutions. Later generations inherited those solutions and added new ones. Rules accumulated, organizations expanded, procedures became established, and exceptions multiplied. Over time, the resulting structures became too complicated for any individual to understand completely.

This is not necessarily evidence of incompetence. Incremental development is often the only practical way for institutions to evolve. But it creates a fundamental limitation: a collection of individually reasonable decisions can produce an unreasonable system. Interoperability offers a way to address that limitation. It allows us to identify the components of a system, describe them using shared concepts, understand their relationships, and evaluate their combined effects. Artificial intelligence greatly expands our ability to perform that work across enormous amounts of information.

The objective is not to eliminate human judgment or replace political and moral choices with mathematical formulas. It is to give human beings a much better understanding of the consequences of their choices. In 1826, a London court could sentence someone to death for a qualifying theft while another court imposed three years for attempting to rape a child. Both courts could be following the law. That was precisely the problem. The individual rules could operate as written while the larger system failed to make sense.

Two hundred years later, we have the opportunity to build systems that do more than accumulate information and enforce rules. We can build systems that understand how their parts relate to one another, detect contradictions, learn from outcomes, and help people make better decisions.That is the deeper promise of interoperability.