The Device Can Measure a Signal. It Cannot Tell You What the Signal Means.

Sensors can detect breath, movement, pulse, skin conductance, and changes in blood oxygenation. Meaning enters later, through models, context, language, and human judgment.

The number on the screen looks like an answer.

A pulse oximeter reads 94 percent. A wearable reports elevated stress. A brain scan highlights a region. Two participants’ breathing patterns begin to converge. A dashboard labels the room coherent.

Each result may contain useful information. None explains itself.

A device can register a physical change, convert it into data, and compare that data with a rule or model. It may even display a word such as calm, stress, attention, or connection. But the word is not waiting inside the signal. It was placed there through a chain of design choices, prior research, statistical assumptions, thresholds, and labels.

That chain can support a responsible inference. It can also conceal one.

For technologies designed to engage awe, presence, altered perception, felt connection, or questions about consciousness and reality, this distinction is not a technical footnote. It is the line between evidence and suggestion.

Measurement begins with a decision

Before an instrument measures anything, someone must decide what is being measured.

Metrology, the science of measurement, calls the quantity intended to be measured the measurand. The International Vocabulary of Metrology also treats a measurement result as more than a bare value. It includes relevant information about the result, which commonly means its uncertainty. The vocabulary makes the point explicit.

This sounds remote from human experience. It is not.

If a respiration sensor reports 10 breaths per minute, its measurand may be breathing rate during a defined interval. That is a physical quantity. If a system reports that the person is calm, the claim has changed. Calm is not identical to respiratory rate. The system has used the measured quantity as evidence for a broader state.

The difference can be stated in five steps:

A sensor detects a physical event.
Hardware and software convert that event into a signal.
Processing turns the signal into a measurement or feature.
A model relates that feature to a proposed state.
A person or system assigns significance to the state.

Errors can enter at every step. So can valid knowledge. The discipline lies in keeping the steps visible.

The sensor may be reliable while the interpretation is weak. The model may perform well in a laboratory and fail in a crowded room. The physiological inference may be sound while the metaphysical conclusion exceeds what was studied.

No amount of precision at the first step grants automatic authority at the fifth.

Even direct measurements carry conditions

Pulse oximetry shows how much can happen before meaning enters the discussion.

A pulse oximeter uses light to estimate arterial blood oxygen saturation. It is a mature medical technology measuring a comparatively well-defined physiological property. Yet the U.S. Food and Drug Administration warns that readings can be affected by poor circulation, skin pigmentation, skin thickness, skin temperature, tobacco use, and fingernail polish. The agency advises interpreting a reading alongside symptoms and changes over time. Read the FDA guidance.

The instrument does not become useless because its reading has limits. It becomes usable when those limits are understood.

This example matters because many systems that claim to measure human experience face a harder problem. They are not estimating one defined physiological quantity and stopping there. They are moving from multiple body signals to a psychological label, then sometimes from that label to a claim about relationship, consciousness, or reality.

If uncertainty matters when estimating oxygen saturation, it matters more when estimating presence.

Arousal is not an emotion

Electrodermal activity measures changes in the skin’s electrical properties associated with sweat-gland activity. It is widely used in psychophysiology because it can provide information about sympathetic nervous-system activation.

It does not directly identify why activation occurred.

A rise in skin conductance may accompany fear, surprise, effort, anticipation, novelty, pain, or excitement. The signal can help establish that arousal changed under defined conditions. Context, experimental design, comparison conditions, participant report, and other measurements are needed to support a more specific explanation.

This is why the Society for Psychophysiological Research’s publication recommendations for electrodermal measurements devote extensive attention to recording methods, environmental conditions, electrodes, signal treatment, response definition, and reporting. The signal must be produced and interpreted through a controlled method. Read the recommendations.

Heart-rate variability presents the same problem in another form. HRV describes variation in the time between heartbeats. Depending on the metric and conditions, it can provide information about autonomic regulation. It is affected by respiration, posture, movement, time of day, substances, health status, age, recording length, and analytical choices.

Researchers therefore recommend controlling or recording these factors and distinguishing among resting, reactive, and recovery measurements. The methodological guidance is a reminder that HRV is not one universal score with one universal meaning.

A system may use HRV as one indicator in a model of stress or regulation. That can be defensible. Saying that the device measured stress, measured safety, or measured spiritual openness erases the inferential work between the heartbeat and the label.

The body supplies evidence. It does not supply the caption.

Brain images do not escape the problem

Brain imaging often carries a stronger cultural authority than peripheral sensors. The image appears to show thought itself.

Functional magnetic resonance imaging does not photograph an idea, emotion, or belief. The common blood-oxygen-level-dependent signal reflects changes involving blood flow and deoxygenated hemoglobin associated with neural activity. The relationship is scientifically valuable and physiologically complex. A review of the BOLD signal describes the vascular and metabolic processes involved.

Interpretation adds another layer. In 2006, cognitive neuroscientist Russell Poldrack examined the logic of inferring a specific mental process from activation in a brain region. The inference is weak when the region is active across many different processes. It becomes stronger only when prior evidence shows that the observed activity is selective for the process being proposed. Read the paper.

This problem is known as reverse inference. Its lesson reaches far beyond fMRI.

When a signal occurs during an experience, the signal does not automatically identify the experience. A correlated change may help test an explanation. It may help discriminate among alternatives. It may become predictive when a model is trained and validated for that purpose. It does not make every interpretation of the state true.

Neural measurement can establish that a brain-related change accompanied a participant’s report of unity. It cannot establish, from that association alone, that unity is nothing but the measured brain process. It also cannot establish that the participant entered a shared field, contacted an external intelligence, or perceived a deeper layer of reality.

The scan does not settle the metaphysics in either direction.

Prediction is not explanation

Modern systems complicate the language because a model can be trained to classify states from patterns too complex for a person to read.

Suppose a wearable combines heart rate, HRV, skin conductance, temperature, and movement. Researchers collect data while participants complete tasks labeled baseline, stress, and amusement. A model learns to distinguish those labeled conditions. If it performs well on people and settings outside the training data, the system may have practical value as a classifier.

But what has it learned?

It has learned a relationship between input patterns and the labels used in the study. Those labels depend on the task, population, measurement protocol, and definition of success. A high classification score does not prove that stress has one physiological signature. It does not establish the cause of the state in a new person. It does not determine what the experience means to that person.

The output may say stress. The underlying claim is narrower: this pattern resembles patterns assigned that label under specified conditions.

That sentence is less impressive. It is also more accurate.

The distinction becomes urgent when systems adapt in real time. A room may change sound or light because respiration slowed. An interface may intensify because arousal rose. A facilitator may tell participants that synchrony indicates connection. The system is no longer observing alone. Its interpretation is changing the environment that produces the next signal.

Measurement, feedback, expectation, and experience can form a loop.

In that loop, a label can become part of the event it claims only to describe.

Synchrony does not contain its own cause

Two signals can become more similar over time. Researchers can quantify alignment in breathing, movement, heart rhythm, speech timing, or other behavior. The result may be meaningful.

It is not self-explanatory.

Participants may synchronize because they hear the same music, follow the same visual pacing, respond to a facilitator, imitate one another, share an emotional event, or influence each other through ordinary sensory channels. Several processes may operate at once. The measured alignment may also depend on the time window, preprocessing choices, and comparison method.

Calling the result coherence does not prove a cause. Calling it connection does not establish the nature of the relationship. Calling it a field effect does not demonstrate that a field exists as a causal mechanism.

Those may be hypotheses worth testing. They remain hypotheses until evidence separates them from alternatives.

This boundary protects inquiry. It prevents a metaphysical interpretation from gaining scientific authority through a sensor display. It also prevents a physiological account from claiming that a participant’s relational, cultural, or spiritual meaning has been fully explained by the detected mechanism.

The interface holds hidden authority

People rarely encounter raw data. They encounter a display.

Color, motion, thresholds, labels, alerts, scores, and summaries tell users what deserves attention. A green ring may imply safety. A rising number may imply progress. A symmetrical pattern may imply harmony. None of these meanings is inherent in the voltage, interval, waveform, or pixel.

The interface is an argument about the data.

That argument may be responsible when it identifies the measured quantity, states how the inference was made, shows uncertainty, distinguishes individual from group results, and avoids claims the system has not validated.

It becomes manipulative when interpretation is presented as detection. A device that reports your energy is blocked or the group has entered collective consciousness is not merely displaying a signal. It is making a claim about a person or reality. The burden of evidence belongs to that claim, not to the sophistication of the hardware.

More sensors do not remove the burden. A multimodal system can improve discrimination while multiplying assumptions. Artificial intelligence can reveal patterns while obscuring how a label was produced. Real-time responsiveness can deepen an experience while making cause harder to isolate.

Technical complexity is not epistemic permission.

What responsible evidence sounds like

The difference between measurement and meaning can be preserved in ordinary language.

Measured: The respiration sensor estimated a decline from 16 to 10 breaths per minute during the session.

Observed: Participants’ breathing rates became more similar during the paced audiovisual sequence.

Reported: Four participants described feeling calmer and more connected afterward.

Inferred: The pattern is consistent with increased physiological regulation under the conditions used.

Not established: The measurements do not identify the cause of the reported connection or demonstrate a metaphysical mechanism.

Each sentence contributes something. None is weakened by the presence of the others.

This is the evidence discipline required for Metaphysical Technology: designed technologies, environments, methods, and frameworks that deliberately engage questions, experiences, or relationships traditionally treated as metaphysical and bring them into operational, observable, testable, or governed form.

The category does not gain credibility by making the device sound omniscient. It gains credibility by identifying what each layer of evidence can support.

A sensor can establish that a signal was detected within stated limits. A study can establish an association under defined conditions. A validated model can support a bounded prediction. A participant can authoritatively report an experience. None of these sources, alone, establishes the ultimate nature or meaning of what occurred.

The number is the beginning

The device can measure a signal. It cannot tell you what the signal means without an interpretive system built by people.

That does not make meaning arbitrary. Interpretations can be tested. Models can be compared. Confounds can be controlled. Predictions can fail. Mechanisms can gain support. Participant accounts can be examined alongside behavior and physiology. Evidence can narrow the possibilities.

Meaning still has to be argued, attributed, and earned.

The responsible question is not whether to trust the number or trust the experience.

It is: What was measured, how was it transformed, what inference followed, which alternatives remain, and who has the authority to name the result?

A device can sharpen that investigation.

It cannot replace it.

Sources

Joint Committee for Guides in Metrology, “International Vocabulary of Metrology: Measurement Result,” VIM3, entry 2.9.
U.S. Food and Drug Administration, “Pulse Oximeter Basics,” March 26, 2025.
Wolfram Boucsein et al., “Publication Recommendations for Electrodermal Measurements,” Psychophysiology 49, no. 8 (2012): 1017–1034.
Sylvain Laborde, Emma Mosley, and Julian F. Thayer, “Heart Rate Variability and Cardiac Vagal Tone in Psychophysiological Research: Recommendations for Experiment Planning, Data Analysis, and Data Reporting,” Frontiers in Psychology 8 (2017).
Elizabeth M. C. Hillman, “Coupling Mechanism and Significance of the BOLD Signal: A Status Report,” Annual Review of Neuroscience 37 (2014): 161–181.
Russell A. Poldrack, “Can Cognitive Processes Be Inferred from Neuroimaging Data?,” Trends in Cognitive Sciences 10, no. 2 (2006): 59–63.