The second instrument produces numbers.
This is its primary advantage over the first, which produces vibration in the bones of my hands, my wrists, my forearms — a reading that corresponds to what the building is doing but which cannot be separated from the fact that I am the one reading it. The piezoelectric array sits on the floor of the corridor at the coordinate, six transducers in a hexagonal pattern, each converting mechanical stress into voltage. The voltages appear on my screen as waveforms. The waveforms do not require my body to exist.
I set it up this morning at 6:40, before the corridor filled. The baseline data was immediate and unambiguous.
Three elements in the 23-31 Hz range, consistent with what the haptic instrument has shown me since October, consistent with the building sensor grid cycling through its monitoring routines in the pattern the nalparam infrastructure maintains even after the LIA lapse dissolved the licensing framework that originally governed it. One cluster around 18 Hz that I have not detected before. The 18 Hz cluster is intermittent: it appears for eleven seconds, disappears for between forty and ninety seconds, reappears. It does not correlate with any building system cycle I can identify. It does not correlate with foot traffic, with the junction box switching patterns, with the thermal sensor recalibration that happens every morning between 5:30 and 7:00 as the grid adjusts to the overnight temperature differential.
The 18 Hz cluster may be new. Or it may have been present since September and I could not detect it with the haptic instrument because 18 Hz is below the threshold of what my hands can reliably distinguish from ambient building vibration. The piezoelectric array has no such limitation. It reads what is there.
This is the advantage. This is also the problem I did not anticipate.
Chae-Gyeol told me at noon.
She stood at the coordinate this morning without the instrument — my instrument, the haptic one, the one that requires her to know she is being measured because it requires me to be present, holding it, reading through my body what the building is doing while she stands at the convergence point and reports what she experiences. She stood there alone for twenty-three minutes. She experienced twenty-three minutes.
No duration compression. No temporal displacement. No sense that the place was doing anything other than being a point in an administratively absent corridor.
I have been sitting with this information for three hours. The studio is quiet except for the building's low hum — the nalparam infrastructure running its cycles, the sensor grid logging ambient conditions that no authorized system is receiving.
The study has documented duration compression at the coordinate across fourteen separate sessions. In each session, the same conditions obtained: the haptic instrument was present, I was present, Chae-Gyeol knew the protocol was active, Bok was photographing from a fixed position at the south end. The compression was consistent — reported durations between 40% and 65% shorter than actual elapsed time. Chae-Gyeol would stand for two hours and report that perhaps forty minutes had passed. The instrument readings correlated: the 23-31 Hz convergence pattern intensified during the periods she later identified as compressed.
But correlation is not causation. And Chae-Gyeol was never naive.
The first visit — September, before I built the instrument, before the study existed as a study — she describes as also compressed. But that description has been refined over five months of repetition, of being asked about it, being recorded describing it, reading Bok's notes about it. The September memory is no longer a memory. It is a reconstruction built from the study's need for an origin point, and Chae-Gyeol knows this, which is why she went back alone this morning.
She went back alone and nothing happened. The coordinate was a point in a corridor.
The design problem is this: I need a naive subject.
Not naive in the sense of uninformed — naive in the experimental sense. A person who stands at the coordinate without knowing what the coordinate is supposed to do, without knowing the study exists, without the accumulated attentional context that Chae-Gyeol and Bok and I have constructed over six months of shared work. A person whose temporal experience at the coordinate is uncontaminated by expectation.
The woman at the south end.
Bok has been tracking her for two months. She appears at the approach threshold — the point where the building's sensor grid shifts from passive monitoring to its denser active-zone configuration, the transition between the corridor's background section and the section where the resonance patterns converge. She stands there. She holds something small — a recording device or a note card, the distance has never been close enough for certainty. She stays for between forty-five seconds and three minutes. She leaves. She has been there nine times in the past eight weeks. She is not random. She has a practice.
Bok found a chalk mark this morning on the exterior wall, below the faded LIA reassignment notice: a time (6:40) and an arrow pointing into the corridor. The mark is not ours. The mark is hers.
She is running a study. A different study, with different instruments or no instruments, with different questions. But she is attending to the same corridor, or being attended to by it, and she does not know about our work, and we do not know about hers.
She is the fourth condition.
The first condition: Chae-Gyeol with the haptic instrument, protocol running, Bok photographing. Duration compression documented across fourteen sessions. The second condition: Chae-Gyeol with the piezoelectric array, protocol running, Bok absent. Not yet tested — scheduled for next week, though Chae-Gyeol's null result may require redesigning the protocol before we proceed. The third condition: Chae-Gyeol alone, no instrument, no protocol. Duration felt like duration. Twenty-three minutes of twenty-three minutes. The fourth condition: the woman. A subject who arrived at the coordinate before we did, who has her own relationship to the corridor, who does not know what we expect to find there.
If she experiences duration compression, the phenomenon exists independent of our measurement context and the study is documenting something real about the building.
If she does not — if her experience at the coordinate is ordinary, unremarkable, just a place she stands in for her own reasons — then the phenomenon is us. The duration compression is not a property of the coordinate or the building's sensor convergence. It is a property of the study itself.
I do not know how to ask her.
The study requires her naivety. The moment I explain what we are investigating, I contaminate the very quality that makes her essential. I cannot mention duration compression without priming her to experience it or to watch for it. I cannot describe the coordinate's suspected properties without making her aware that the coordinate has properties worth suspecting. The act of recruiting her as a subject destroys the condition I need her to fulfill.
But I also cannot measure her perception without her knowledge. The piezoelectric array is passive — it reads building vibration, not human temporal experience. To know whether she experiences duration compression, I need her to report what time felt like. And to report what time felt like, she needs to know that someone is interested in what time felt like. And to know that someone is interested in what time felt like at this specific location is to know that time at this location might behave in a way worth investigating.
The CIRB framework — the Coupling-Informed Research Board that governs lending research in the Lend District — has protocols for naive-subject studies, but those protocols were designed for coupling research, where the variable is neurological and the measurement is physiological. The CIRB can authorize a study where the subject does not know which session involves active lending and which is control. What I am trying to measure is phenomenological. There is no biomarker for duration compression. There is only the report of the person standing in the place, and the report requires awareness, and awareness requires consent, and consent requires disclosure that defeats the purpose of the measurement.
Bok said: ask her what she is doing here. Start with her study, not ours. Find out what she already knows about the corridor before you tell her anything.
This is probably right. Bok's instinct for the social architecture of the corridor is better than mine — she has been here longer, she understands the etiquette of shared space in nalparam-conditioned buildings. The way people who use a space that exists in administrative absence develop informal protocols for coexistence, acknowledge each other's presence without naming what they are doing there. The woman does not need my permission to be at the coordinate. I do not need hers. But the moment we acknowledge each other's presence as researchers rather than as passersby, the corridor's two studies become one study, and neither of us can go back to the condition of not knowing what the other is looking for.
I will ask Bok to introduce us. Bok has been trying to learn her name for two months. This is a reason to succeed.
The piezoelectric array is still running at the coordinate. It has been recording since 6:40 this morning. By now it has accumulated eight hours of continuous data — the 23-31 Hz baseline, the intermittent 18 Hz cluster, and everything the corridor's nalparam infrastructure has produced today while I have been sitting in Bok's studio, not at the coordinate, thinking about a woman I have not spoken to and a design problem I cannot solve without solving the ethics of observation first.
The instrument does not share this difficulty. It reads what is there. The numbers accumulate without interpretation, without the need for consent, without awareness that they are data. The building speaks its frequencies into the transducers, the transducers convert mechanical stress into voltage, the voltage becomes numbers on a screen. No one needs to agree to any of this.
I will go back to the coordinate this evening and retrieve the array. I will look at eight hours of waveform data and I will not know, from the numbers alone, whether the corridor did anything remarkable while I was gone. The numbers will show patterns. The patterns will correlate with building cycles, foot traffic, thermal variation, and one intermittent 18 Hz signal whose source I have not yet identified.
The numbers will not tell me whether time behaved differently at the coordinate today. Only a person standing there could tell me that, and the person has to not know I am asking.