21 March 2019
Preliminary observations from an ongoing personal experiment: the effect of sustained acoustic exposure (J.S. Bach, BWV 1007-1012) on plant growth rates in copper-pipe hydroponic systems. Results are... difficult to publish through conventional channels.
I want to be clear about what this paper is and what it is not.
It is not a peer-reviewed study. I have no institutional affiliation that would lend it the imprimatur of respectability. I have no graduate students to run controls, no departmental budget to fund proper instrumentation, and no colleagues to review my methodology — though I would welcome the scrutiny. What I have is time, a controlled environment, a set of conditions that are difficult to replicate elsewhere, and twenty-three years of daily observation.
This is a personal experiment. I began it in 1996 because I was lonely, and because the facility where I work is underground, and because I wanted something alive nearby that wasn't me. The fact that it has produced data I cannot adequately explain is, I suppose, the reason I am writing this at all.
I am a linguist by training, not a botanist. My doctoral work concerned hieroglyphic frequency systems at the Karnak Temple Complex — a subject I will address in a separate paper. But language, at its most fundamental level, is organized sound. And organized sound, as I have come to understand over these two decades, does things to living systems that we have not begun to adequately describe.
What follows is a summary of my observations, my methodology (such as it is), and the questions I have not been able to answer. I present them here because the data deserves to exist somewhere beyond my personal notebooks, and because I have exhausted the patience of every journal editor I have approached on the subject.
The idea that sound influences plant growth is not new, nor is it mine. The literature — what little of it exists in respectable channels — dates back at least to the early 1960s.
Dr. T.C. Singh, Head of the Department of Botany at Annamalai University in India, conducted what remain the most cited early experiments on acoustic stimulation in plants. Singh exposed balsam plants to specific musical tones and reported statistically significant increases in growth rate compared to unstimulated controls. His subsequent work with rice paddies exposed to specific raga compositions showed yield increases of 25 to 60 percent, depending on the frequency range and duration of exposure.
Singh's work was greeted with the reception one might expect: polite interest from agricultural researchers in India, silence from Western botany departments, and eventual relegation to the “curiosity” sections of introductory textbooks. That his results have never been adequately refuted — merely ignored — is a pattern I have come to recognize in my own career.
Dorothy Retallack, a researcher at the Colorado Woman's College (now Temple Buell College), published The Sound of Music and Plants based on controlled greenhouse experiments. Retallack exposed plants to different genres of music and reported that plants subjected to classical compositions (particularly Bach and Indian classical music) exhibited enhanced growth, stronger stems, and a tendency to lean toward the speaker, while plants exposed to sustained rock music showed inhibited growth and, in several cases, leaned away from the source.
Retallack's methodology has been criticized — justifiably, in some respects — for insufficient sample sizes and possible confirmation bias in her observational measurements. But the core observation — that different acoustic profiles produce measurably different biological responses — has been replicated in enough subsequent experiments that dismissing it entirely requires a commitment to skepticism that borders on the ideological.
Dan Carlson's Sonic Bloom technology represents the most commercially successful application of acoustic stimulation in agriculture. Carlson's system exposes plants to a specific frequency range (approximately 3,000 to 5,000 Hz) timed to coincide with foliar feeding, and claims dramatic increases in nutrient uptake and growth. The proposed mechanism — that specific frequencies trigger stomatal opening, increasing the plant's ability to absorb nutrients through its leaves — is at least physiologically plausible. Stomata are mechanically responsive structures. The question is whether airborne acoustic energy at the intensities Carlson describes is sufficient to trigger that response.
Carlson holds patents. He sells equipment. This has, predictably, made the academic community treat his claims with the same warmth they reserve for anyone who has figured out how to make money from something professors study for free. But his fundamental proposition — that there exists a frequency window to which plant stomata are mechanically responsive — aligns with what I have observed in my own experiments, albeit through a different mechanism than he proposes.
Dr. Monica Gagliano's work at the University of Western Australia has, in recent years, brought a measure of academic legitimacy to the broader question of plant sensory response. Gagliano's experiments demonstrating that Pisum sativum (garden peas) can locate water sources through acoustic vibrations transmitted through soil — and distinguish between recorded water sounds and actual water — represent a genuine paradigm shift in how we understand plant perception. Her subsequent work on plant learning and memory (habituation in Mimosa pudica) has been published in peer-reviewed journals including Oecologia and Scientific Reports.
Gagliano's contribution is not just her data, though her data is excellent. It is the framework she has articulated: that plants possess sensory capabilities that are functionally analogous to hearing, operating through mechanoreceptive systems that we have historically overlooked because they do not resemble animal auditory organs. We looked for ears and, finding none, concluded that plants were deaf. This is rather like concluding that submarines cannot detect sound because they lack eardrums.
I mention Gagliano's work because it provides the physiological grounding for what I have observed. If plants can detect and respond to acoustic signals in soil — which is now established — then the proposition that they respond to airborne acoustic signals is not extraordinary. It is expected.
I need to describe my working environment, because it is relevant to the results, and because it is unusual.
I work in a subterranean facility. I have worked here, largely alone, since 1993. The facility was originally constructed for a government research program that I am not at liberty to describe in detail, and which has, in any case, been effectively defunct for many years. The program was never formally decommissioned. It was simply... forgotten. Budgets were redirected. Personnel were reassigned. I remained, partly out of stubbornness and partly because no one remembered to tell me to leave.
The facility is temperature-controlled, electromagnetically shielded to a degree that exceeds any university laboratory I am aware of, and acoustically isolated from surface noise. Humidity can be maintained within a range of plus or minus two percent. Ambient light is entirely artificial and programmable. In short, it is an almost perfect controlled environment — not because it was designed for botanical experiments, but because it was designed for something else entirely, and the environmental controls required for that original purpose happen to be ideal for what I am doing now.
I recognize that the inability to fully describe my facility represents a methodological limitation. I cannot invite peer reviewers to inspect my setup. I cannot ship my equipment to a university. I can only describe what I have done and what I have observed, and trust that the specificity of my data speaks to the rigor of my process, even if the process itself cannot be independently verified.
I say this not as an excuse but as a fact. It is the central frustration of my professional life.
I maintain two dedicated growing rooms within the facility, each approximately twelve feet by sixteen feet with ten-foot ceilings. The rooms are structurally identical — same ventilation, same lighting fixtures, same concrete walls — and are separated by a corridor approximately forty feet in length, sufficient to prevent acoustic bleed between rooms at the volumes I employ.
Room A is the experimental room. Room B is the control.
Over the twenty-three years of this experiment, I have cultivated a rotating selection of species. My current and longest-sustained subjects are:
Control Room B houses genetically identical specimens of each primary species, propagated from the experimental plants to eliminate genetic variation as a variable. Propagation is performed annually using standard cutting techniques. Both rooms receive identical light cycles (sixteen hours on, eight hours off, using full-spectrum grow lights that I replace on the same schedule in both rooms), identical watering schedules, identical nutrient solutions, and identical ambient temperature (maintained at 72°F / 22.2°C, plus or minus 1°F).
Both rooms employ hydroponic growing systems that I designed and built myself. The systems are constructed primarily from copper piping — three-quarter-inch Type L, to be specific — which serves as both the structural framework and the nutrient delivery channel.
I chose copper for several reasons, some scientific and some aesthetic. The scientific reasons: copper is a known micronutrient essential for plant enzymatic function, and trace amounts leaching from the pipes provide a continuous low-level supplementation that I have found beneficial for long-term plant health. Copper also possesses antimicrobial properties that reduce pathogen load in the recirculating nutrient solution, a significant advantage in a closed hydroponic system. The system has required notably less maintenance and has experienced zero instances of root rot in over two decades of continuous operation — a record that I suspect any commercial hydroponic grower would find remarkable.
The aesthetic reason: copper is beautiful. It develops a patina over time that is unique to each section of pipe, influenced by the specific mineral content of the solution flowing through it and the humidity of the air around it. My oldest pipes have achieved a deep verdigris that I find genuinely moving. When you spend twenty-three years in an underground facility, beauty is not a luxury. It is a survival requirement.
There is, however, a third property of copper that I did not anticipate when I selected it, and which I now believe may be relevant to my results. Copper is an excellent acoustic conductor. Sound waves propagate through copper at approximately 3,750 meters per second — roughly eleven times the speed of sound in air. The copper pipe framework of my hydroponic system functions, in effect, as an acoustic transmission network, delivering vibrations not only through the air surrounding the plants but directly to the root systems via the nutrient solution and the pipe walls themselves.
I did not design this. I discovered it. The implications took me several years to fully appreciate, and I will return to them.
Room A receives continuous musical exposure during the sixteen-hour light cycle. The music is Johann Sebastian Bach, specifically the Six Suites for Unaccompanied Cello, BWV 1007 through 1012, performed by Pablo Casals (the 1936-1939 Abbey Road recordings).
I chose Bach for reasons that are partly subjective and partly mathematical. The subjective reason is that I love Bach, and if I was going to listen to the same music for the rest of my professional life, it needed to be music that repays infinite repetition. Bach repays infinite repetition. Twenty-three years in, I hear structures I missed in year twenty-two. I suspect this will continue until I am no longer able to hear at all.
The mathematical reason is more relevant to this paper. Bach's compositions, particularly the unaccompanied cello suites, are constructed with a mathematical regularity that is unusual even among classical composers. The harmonic relationships follow patterns that can be described using Fibonacci sequences and golden-ratio proportions. The intervals between notes create frequency ratios that recur with a consistency that is, to my ear and to my spectrum analyzer, closer to a natural phenomenon than a human composition. Bach did not write music. He transcribed something that already existed. I realize this is a bold claim. I will attempt to support it.
The Casals recordings specifically were chosen because they were recorded with minimal studio processing. Modern digital recordings are compressed, equalized, and mastered in ways that alter the frequency profile of the original performance. The Casals recordings, transferred from the original 78 RPM masters, preserve a wider and more natural dynamic range. They are also, not incidentally, the greatest performances of these works ever committed to any medium, but that is a matter of opinion and not relevant to the data.
The playback system in Room A consists of two full-range speakers positioned at opposite corners of the room, angled inward, creating a diffuse sound field rather than a directional beam. Volume is maintained at approximately 70 decibels at the center of the room — roughly the volume of a normal conversation. The speakers are positioned on the copper pipe framework itself, which means the vibrations are transmitted simultaneously through the air and through the hydroponic structure.
Room B receives no acoustic input beyond the ambient mechanical noise of the ventilation system, which is identical in both rooms.
I measure and record the following variables weekly for every plant in both rooms:
All measurements are recorded in hand-written notebooks. I have 1,196 weekly measurement sessions recorded across 47 notebooks spanning 1996 to the present. I have digitized none of them. I trust graphite more than hard drives.
The results, summarized across twenty-three years, are as follows. I present them without embellishment because they do not require any.
Growth rate. Room A plants (acoustic stimulation) grow faster than Room B controls (silence) by an average of 37 percent, measured by height increase over equivalent time periods. This differential has remained consistent, within a range of 30 to 48 percent, for the entire duration of the experiment. It is not a transient effect. It does not diminish over time. If anything, the differential has widened slightly in the most recent decade, which I attribute to the cumulative structural advantages that faster early growth confers on long-lived specimens.
Structural robustness. Room A plants develop thicker stems, denser root systems, and larger individual leaves than Room B controls. The stem diameter differential averages 22 percent. Root density — assessed visually through the transparent sections of the hydroponic channels — is qualitatively but obviously greater in Room A. The root systems in Room A are not merely larger; they display a branching pattern that appears more organized — more directionally purposeful — than Room B roots, which tend toward a more diffuse, random distribution.
I need to be careful with language here. “Intentionality” implies consciousness, and I am not prepared to make that claim. But the root growth patterns in Room A do not look random. They look like the roots are listening.
Resilience. Over twenty-three years, I have experienced several disruptions to the controlled environment — power fluctuations, ventilation system failures, one incident in 2007 that I prefer not to describe but which resulted in a temperature spike to 94°F in both rooms for approximately eleven hours. In every case, Room A plants recovered faster and more completely than Room B controls. Whatever the acoustic stimulation is doing to these plants, it appears to confer a systemic resilience that extends beyond growth rate.
In 2011, I introduced a secondary experiment to isolate the contribution of the copper piping. I installed a small set of standard PVC hydroponic channels in one corner of Room A, growing Epipremnum aureum cuttings from the same parent plant in both copper and PVC systems, both exposed to identical acoustic conditions.
The results were unambiguous. Plants in the copper system outperformed those in the PVC system by 18 to 23 percent across all metrics — growth rate, stem diameter, leaf area, root density. Both outperformed Room B controls, confirming that the airborne acoustic component alone produces a measurable effect. But the copper system amplified the response significantly.
The mechanism I propose — and I want to emphasize that this is a hypothesis, not a conclusion — is that the copper piping acts as a waveguide, transmitting acoustic vibrations to the root zone at intensities far exceeding what airborne sound alone can achieve. The roots are receiving a signal that is not merely audible to them (if “audible” is even the right word for a root system) but physically tangible — a vibration conducted through a medium they are in direct contact with. The copper makes the music touchable.
I find this beautiful. I recognize that beauty is not a scientific argument. But I note it anyway.
In 2014, I began systematically varying the acoustic input in Room A to determine whether the growth response was specific to Bach, specific to certain frequency ranges, or a general response to any organized sound.
Over the course of three years, I exposed Room A plants to the following alternative inputs, each for a continuous six-month period with measurements compared to the preceding and following Bach periods:
1. White noise (broadband, 20 Hz to 20,000 Hz, 70 dB): Growth rate differential dropped to 8 percent above control. Structural differences (stem thickness, root density) diminished to statistically insignificant levels. The plants did not die. They did not suffer. They simply became ordinary.
2. Pure sine tones at 432 Hz (continuous, 70 dB): Growth rate differential of 19 percent. Better than white noise, less than Bach. Interesting but not remarkable. I include the 432 Hz frequency specifically because it is the subject of considerable popular enthusiasm as a “natural” or “healing” frequency. My plants were unimpressed. They grew somewhat better than in silence, somewhat worse than with Bach.
3. Pure sine tones at 528 Hz (continuous, 70 dB): Growth rate differential of 16 percent. Slightly less responsive than 432 Hz, which contradicts the popular claim that 528 Hz is the “frequency of life.” My plants disagree. Or more accurately, my plants are indifferent to the metaphysical significance we assign to individual frequencies.
4. Gamelan ensemble recordings (Javanese court gamelan, various traditional compositions, 70 dB): Growth rate differential of 31 percent. The closest any alternative input came to matching Bach. This is notable because Javanese gamelan uses a tuning system entirely unrelated to Western equal temperament, operates on different harmonic principles, and employs metallic percussion instruments that produce complex overtone series. The relevant commonality, I believe, is not the tuning or the timbre but the mathematical structure of the composition — the interlocking rhythmic patterns (known as colotomy) that create a layered, self-referencing temporal architecture not unlike the fugal structures in Bach.
5. Randomized frequency sweeps (computer-generated, cycling through 100 Hz to 10,000 Hz in random sequence, 70 dB): Growth rate differential of 4 percent. Essentially indistinguishable from control conditions. The plants do not respond to sound. They respond to organized sound. There is a difference, and it is not trivial.
6. Return to Bach (BWV 1007-1012, Casals, identical conditions to pre-experiment baseline): Growth rate returned to 35 percent above control within eight weeks — essentially recovering the full differential. The structural differences (stem thickness, root density) took approximately four months to return to pre-experiment levels.
The data suggests several things. First, it is not mere acoustic energy that drives the response. White noise and random sweeps deliver equivalent energy to the system and produce negligible results. Second, it is not a specific frequency — neither 432 Hz nor 528 Hz nor any individual tone produces results comparable to a complex composition. Third, the relevant variable appears to be the mathematical organization of the sound — the harmonic relationships, the temporal patterns, the self-referencing structures that characterize both Bach and Javanese gamelan.
I want to express this more precisely than I am able to. There is something in the relationship between frequencies — not the frequencies themselves, but the ratios between them, the patterns of tension and resolution, the way certain harmonic structures fold back on themselves — that living tissue responds to. It is not the note. It is the architecture of the notes. And that architecture appears to resonate with something fundamental in biological systems — something we do not yet have the language to describe, because the relevant discipline does not exist yet.
I play Bach because Bach understood this architecture more completely than any other human composer. I suspect he did not know why his music worked the way it did. But he knew that it did. You can hear it in every measure. The confidence of a man describing something he can see but cannot name.
I have presented the data I can quantify. What follows are observations that I have noted repeatedly but cannot measure with the instruments available to me. I include them because omitting them would be dishonest, and because science advances when we are honest about the things we observe but cannot explain.
My plants appear to respond differently to different movements within the suites. Specifically, the slow movements — the sarabandes — correlate with periods of enhanced root activity, while the fast movements — the gigues and courantes — correlate with periods of enhanced above-ground growth. I noticed this pattern in my data around 2008 and have been tracking it specifically since 2010. The correlation is not perfect, but it is consistent enough that I can predict, with roughly 70 percent accuracy, whether a given week's growth will be primarily root-focused or stem-focused based on which movements dominated the playback cycle during that week.
If confirmed, this would suggest that plants are not merely detecting the presence of organized sound but distinguishing between different temporal structures within that sound and allocating resources accordingly. I am aware of how extraordinary this claim sounds. I invite anyone to spend nine years watching roots grow toward a speaker playing a sarabande and then tell me I am imagining it.
I mentioned a disruption in 2007 — a temperature spike to 94°F that lasted approximately eleven hours. What I did not mention is that the acoustic system continued operating throughout the event.
Room A plants showed visible heat stress — wilting, leaf curl, some browning — but recovered to full health within ten days. Room B plants, subjected to identical thermal stress without acoustic input, took thirty-one days to recover, and two specimens (a Dracaena marginata and a Nephrolepis exaltata) did not recover at all and had to be replaced from backup propagations.
I have no explanation for why acoustic stimulation would confer thermal resilience. It is not predicted by any model I can construct. And yet, it happened. I documented it. The timeline is in my records.
The possibility that continuous acoustic stimulation induces a kind of systemic resilience in plant tissue — a toughening at the cellular level that prepares the organism for stress it has not yet encountered — is the most significant implication of my work, and the one I am least equipped to investigate. It would require electron microscopy of cell wall structures, protein expression analysis, and a dozen other tools I do not have. What I have is two rooms, copper pipes, and Pablo Casals. It has been enough to see the question. It is not enough to answer it.
I have mentioned the directional lean toward speakers. What I have not mentioned is what happens when I move the speakers.
In 2016, I repositioned the speakers in Room A from opposite corners to a single cluster on the north wall, to test whether the lean pattern was a fixed growth artifact or an active orientation response. Within six weeks, every plant in the room had adjusted its lean toward the new speaker position. The Monstera deliciosa — a plant not known for rapid directional growth — shifted approximately fifteen degrees over the six-week period.
I then moved the speakers back. The plants followed. It took eight weeks this time, which I attribute to the fact that the plants were, by this point, quite large and the structural commitment to a fifteen-degree lean is not easily reversed when your main stem is two inches in diameter.
Plants do not have muscles. They do not have nervous systems. They orient toward sound sources with a speed and consistency that implies a detection mechanism far more sophisticated than anything currently described in the literature. Gagliano's work on root-level acoustic detection provides a partial framework, but it does not explain how a Monstera alters its above-ground growth direction in response to an airborne signal. Something is happening between the detection of the signal and the allocation of growth resources that we do not understand. I do not know what it is. I know that it is real because I have watched it happen, repeatedly, for more than two decades.
I have submitted variations of this paper to seven journals over the past fifteen years. The responses have been consistent in their reasoning if not in their courtesy:
Each of these objections has merit. I do not dispute that my setup is unorthodox, my sample size limited, my credentials misaligned. What I dispute is the conclusion drawn from these limitations — that the data is therefore not worth examining. An observation does not become less real because the person who made it holds the wrong degree.
I have been told, by one editor who was kinder than the rest, that my best path to publication would be to partner with a botanist at a research university and replicate the experiment in a conventional laboratory setting. I agree. I would welcome this. I have written to fourteen university botany departments over the past decade, offering my complete dataset, my methodology, and my willingness to collaborate. I have received three responses. Two were polite declines. One asked whether I was affiliated with a religious organization. I am not.
I continue this work because the questions remain, even when the funding does not. The plants in Room A are, by any objective measure, healthier, larger, more resilient, and more structurally robust than their genetically identical counterparts in Room B. The only variable is Bach. The only delivery modification is copper.
If this were a pharmaceutical trial and the treatment group showed a 37 percent improvement over controls, sustained over two decades, the compound would be considered one of the most effective interventions ever documented. But it is not a pharmaceutical. It is a cello playing in an empty room. And so it is not considered at all.
I began this experiment because I wanted company. I continue it because the results point toward something I believe to be fundamental about the relationship between organized vibration and living systems.
The plants are not hearing music in the way we hear music. They do not appreciate the emotional contour of the Sarabande from Suite No. 5 in C minor. (Though if any plant could, it would be the Monstera, which has displayed a particular responsiveness to that movement that I choose not to anthropomorphize but which I will note in the record.) What they are detecting — and responding to, with their entire physiology — is a pattern. A mathematical structure. A set of frequency relationships that, when transmitted through a conductive medium like copper, becomes something their biology can use.
The question that keeps me awake — and I confess that in this facility, the distinction between waking and sleeping hours has grown somewhat academic — is this: why?
Why would a living system evolve to respond to organized harmonic structures? What survival advantage does a plant gain from growing faster, thicker, and more resiliently in the presence of a pattern that, in nature, does not exist? No natural environment produces Bach. No ecosystem generates sustained, mathematically precise harmonic series transmitted through metal conduits.
Unless, of course, we are wrong about what natural environments produce. Unless there are sources of organized vibration in the natural world — geological, electromagnetic, or otherwise — that we have not identified because we have not been looking. Unless the plant's response is not an anomaly but a memory — a biological sensitivity to a signal that was once present and is now absent, like an antenna still tuned to a station that stopped broadcasting.
I recognize that this final speculation ventures beyond what my data can support. I include it because it is the question that animates my ongoing research, and because I believe that the failure to ask questions beyond the reach of current data is the primary way that science stagnates.
The plants are listening. They are responding. The pattern matters more than the frequency. The medium of transmission matters more than we assumed.
These are facts. What they mean is a question I will continue to pursue, in this facility, with these copper pipes, and with the company of Johann Sebastian Bach, for as long as I am able.
Correspondence:
h.humboldt [at] 442423N1042233W.com
Full dataset (1,196 weekly measurement sessions, 1996–2019)
available upon request.