Content uploaded by Wayne B Jonas
Author content
All content in this area was uploaded by Wayne B Jonas on Sep 09, 2019
Content may be subject to copyright.
The FASEB Journal •Hypothesis
Can specific biological signals be digitized?
Wayne B. Jonas,*
,†,1,2
John A. Ives,*
,‡,1
Florence Rollwagen,
†,1
Daniel W. Denman,
§
Kenneth Hintz,
㥋
Mitchell Hammer,
¶
Cindy Crawford,*
,†,1
and Kurt Henry
**,1
*Samueli Institute for Information Biology, Alexandria, Virginia, 22314, USA
†
Uniformed Services University of the Health Sciences, Bethesda, Maryland, USA;
‡
Walter Reed Army
Institute of Research, Washington, D.C., USA;
§
University of Maryland, College Park, Maryland, USA;
㛳
George Mason University, Fairfax, Virginia, USA;
¶
American University, Washington, D.C., USA; and
**Defense Advance Research Projects Agency, Arlington, Virginia, USA
ABSTRACT At the request of the United States De-
fense Advanced Research Projects Agency, we at-
tempted to replicate the data of Professor Jacques
Benveniste that digital signals recorded on a computer
disc produce specific biological effects. The hypothesis
was that a digitized thrombin inhibitor signal would
inhibit the fibrinogen-thrombin coagulation pathway.
Because of the controversies associated with previous
research of Prof. Benveniste, we developed a system
for the management of social controversy in science
that incorporated an expert in social communication
and conflict management. The social management ap-
proach was an adaptation of interactional communica-
tion theory, for management of areas that interfere
with the conduct of good science. This process allowed
us to successfully complete a coordinated effort by a
multidisciplinary team, including Prof. Benveniste, a
hematologist, engineer, skeptic, statistician, neurosci-
entist and conflict management expert. Our team
found no replicable effects from digital signals.—
Jonas, W. B., Ives, J. A., Rollwagen, F., Denman, D. W.,
Hintz K., Hammer, M., Crawford, C., Henry, K. Can
specific biological signals be digitized? FASEB J. 20,
23–28 (2006)
Key Words: social management of science 䡠controversial sci-
ence claims 䡠electromagnetic signals 䡠digital biology 䡠throm-
bin 䡠fibrinogen 䡠coagulation
BACKGROUND
In 1988, the journal NATURE published an article on a
controversial claim by Professor Jacques Benveniste of
INSERM in France, and others, reporting that ultra-low
doses of IgE could induce degranulation of sensitized
basophils even when the IgE was diluted beyond the
probable presence of the original molecules (1). The
article was accompanied by an editorial expressing
disbelief in the phenomenon (2). In an unorthodox
and controversial move, Nature bypassed the normal
scientific process of waiting for independent replica-
tion and sent a team to investigate the claim in Ben-
veniste’s laboratory. Under emotionally charged condi-
tions, they performed and subsequently published their
results (3). Charge and counter-charge ensued (mostly
in the press) (4) while subsequent attempts to scientif-
ically investigate the claim produced negative (5),
equivocal (6, 7), and positive (8, 9) reports. While the
topic largely disappeared from the scientific literature,
it continues to be discussed and evaluated by the public
in prominent media places such as the BBC and ABC’s
20/20 (http://www.abcnews.go.com/sections/2020/
GiveMeABreak/GMAB_homeopathy_040130-1.html).
During the ensuing years Prof. Benveniste has pur-
sued possible mechanisms for the claimed effect. He
reported in The FASEB Journal and in other journals
that he could mimic the electromagnetic nature of
molecule-receptor interaction with digital signals to
which biological systems respond in a specific manner
(10, 11) (www.digibioSA.com). While it is known that
molecular interactions in biological systems can be
influenced by low-level electromagnetic fields (12), the
ability to specify and deliver signals through electro-
magnetic mechanisms has not otherwise been re-
ported. The implications for digitally mimicked biolog-
ical signals led the Defense Advance Research Projects
Agency (DARPA), a branch of the U.S. Department of
Defense, to solicit us for an independent replication of
Prof. Benveniste’s reports.
MANAGING SOCIAL CONFLICT
The controversial nature of the topic and the handling of
previous research claims of Prof. Benveniste by the scien-
tific community required, in addition to normal scientific
methods of investigation, a social and communication
management process that was capable of dealing with
conflicting interpersonal dynamics among vested parties
1
The views, opinions, and assertions expressed in this
article are those of the authors and do not reflect official
policy of the Department of Defense, the Department of
Health and Human Services, or the U.S. Government. This
research was supported by a grant from the Defense Ad-
vanced Research Projects Agency.
2
Correspondence: Samueli Institute, 1700 Diagonal Rd., Suite
400, Alexandria, VA 22314, USA. E-mail: wjonas@siib.org
doi: 10.1096/fj.05-3815hyp
230892-6638/06/0020-0023 © FASEB
in the research effort. The approach we used represents a
modification of a communication-based process originally
developed within the crisis management arena that fo-
cuses on the functional meaning of discourse taking place
under conditions of disagreement and emotional inten-
sity (13, 14). The process is an adaptation from interac-
tional communication theory and identifies four conflict
discourse “frames” that focus communication among par-
ties around interactional “triggers” of conflict escalation
and de-escalation (15–17). These four frames and their
corresponding behaviors revolve around 1) substantive
disagreements around the goals and methods of the
research effort; 2) attunement issues (e.g., degree of trust
among members of the research team, acceptability of
different levels of power and influence, demonstrated
consideration for one another); 3) face concerns (e.g.,
degree to which member of the research team perceive
their reputation or professional identity is being attacked
or honored); and 4) emotional discord (e.g., degree to
which negative emotions arise that increase mistrust and
inhibit accomplishment of research objectives).
SOCIAL CONFLICT IN MEDICAL RESEARCH
We previously developed a process for the adaptation
and application of this social and communication man-
agement process to controversial scientific research
topics. Specifically, under the direction of Dr. Wayne
Jonas, then director of the Office of Alternative Medi-
cine (OAM) at the National Institutes of Health (NIH),
three major projects were undertaken that provided
the substantive basis for development of a social and
communication management process for resolving con-
flicts that can arise in controversial research topics.
These projects were 1) a comprehensive case study
analysis of the Burzynski antineoplaston clinical trial on
adult brain tumors conducted by the National Cancer
Institute (NCI) (18, 19); 2) a one-day meeting that
focused on the design and implementation of a process
for managing conflict dynamics interpretation and
judgment processes of OAM sponsored research (20);
and 3) a 3-day meeting and follow-up activities that
involved 80 cancer researchers and practitioners from
the complementary and alternative cancer therapy
community and researchers from the NCI. The focus of
this effort was on the development of conflict manage-
ment processes to be integrated into a practice out-
comes monitoring and evaluation system (POMES)
designed to address research questions related to con-
troversial cancer therapies (21).
These findings indicated that difficulties negatively
affecting the conduct of research efforts on controversial
topics not only reflect differences of opinion on substan-
tive issues (e.g., research protocols, appropriate statistical
analysis) but, more important, reflect social conflict
among the parties. This leads to interpretive and judg-
ment biases that result in selective filtering and recall of
data, dismissal of nonconforming data, increased mistrust
of the motives of the other party, reduced ability to
empathize with the other, simplified thinking, and over-
reliance on prior expectations and experiences (18, 19).
INSERTION OF SOCIAL MANAGEMENT INTO A
RESEARCH PROJECT
The most visible aspect of the social management
process into this study involved the integration of a
communication and conflict resolution professional
into the research team. His role was to observe, moni-
tor, and evaluate the social dynamics during implemen-
tation of the research project. This involved several
functions, including 1) intervention into group deci-
sion-making processes around both substantive issues
and assessment of attunement, face, and emotional
discord dimensions; 2) application of conflict commu-
nication skills (e.g., active listening) to key prioritiza-
tion concerns (e.g., how the research goals are set),
development and review of the research approach
(e.g., how the research design and protocols are estab-
lished), execution of the study (e.g., how are protocols
actually implemented, how are changes/modifications
made to the previously agreed upon research proto-
cols), and interpretation and evaluation of the research
findings (e.g., what values and criteria are applied to
judge the quality and validity of the research effort); 3)
conduct of individual interviews, outside of the normal
meetings, with research team members to assess the
quality of the group process; 4), discussion of observa-
tions of the ongoing social and management process
during group meetings; and 5) at the conclusion of the
study, preparation of written analysis along with a
formal presentation to the research team of the social
dynamics that characterized the research effort. This
system, while it had never been formally tested in a
laboratory setting, seemed appropriate to use during an
independent test of Prof. Benventiste’s claims, re-
quested of us by the Department of Defense.
THE DIGITAL BIOLOGY EXPERIMENTS
The attempted replication project was divided into four
phases: protocol development, pre-pilot, pilot, and test
phases. In the Protocol Development phase a detailed
preliminary protocol was developed and submitted to
DARPA for review including a description of the com-
munication and management process to be used. The
protocol was reviewed and agreed to by Prof. Ben-
veniste and his staff, including the criterion that he
would serve as a consultant and not an author of any
final publications that ensued. This was to assure that
the replication attempt and subsequent paper were
independent. This protocol was further developed after
a multidisciplinary consultant team consisting of a
biologist, hematologist, statistician, electrical engineer,
and communication/conflict expert was recruited.
Feedback and modifications of the experimental pro-
24 Vol. 20 January 2006 JONAS ET AL.The FASEB Journal
cedures were documented via meetings, memoranda,
and email throughout the study.
The pre-pilot phase was conducted in our laboratory
July 14–21, 2001 by Didier Guillonnet and Jamal Aissa
from Benveniste’s laboratory. They set up an Automated
Bio-Analyzer (ABA) machine that delivers the digital
signals and automatically runs the experiments (Fig. 1A).
A number of modifications to the ABA program and
protocol were implemented and verified at that time,
including blinding and randomization procedures. These
same two individuals, along with Prof. Benveniste, re-
turned on October 30 through November 3, 2001 to run
the first pilot phase under the modified protocol. The test
phase was then conducted by our research team using the
same protocol, but independent of Prof. Benveniste’s
team. Between the pre-pilot and pilot experiments, a
series of blank runs were made on the ABA with water
signals to estimate sample sizes required to observe a 20%
effect difference between groups with a Pvalue set at 0.05
for hypothesis determination.
According to Prof. Benveniste, a biological system
that can be used to detect digital signals is the inhibi-
tion of fibrinogen-thrombin coagulation by a digitized
thrombin inhibitor (DTI). Our working hypothesis was
that there would be no difference in coagulation rates
between groups exposed to a digitized thrombin inhib-
itor (DTI), a digitized water signal (WAT), and no
signal (NS), as measured by optical density (OD) curves
of samples exposed to these signals.
To maximize objectivity and reliability, the DigiBio
research team set up an ABA robot that performs the
entire protocol, including sample pipetting and deliv-
ery of the signal, with minimal human intervention
(Fig. 1A, B). The protocol can be summarized as
follows. Thrombin dissolved in water (0.3 g/mL) is
exposed to the computer recording of either DTI or
control solutions consisting of either no electromag-
netic (EM) signal or water signal. After exposure, the
samples are mixed with fibrinogen (24 mg/mL in 0.1%
sodium azide, to prevent bacterial growth) and distrib-
uted in 96-well plates. Coagulation was assessed by
spectrophotometry and expressed as OD. The entire
procedure was automated except for data analysis,
stock preparation of fibrinogen and thrombin, and
their loading into ABA reservoirs.
The basic ABA procedure is as follows. After placement
of fibrinogen and thrombin into appropriate reservoirs,
the computer program is started. The machine distributes
the thrombin into Eppendorf tubes and then places the
tube into an EM coil inside of a grounded copper tube
open at the top. The computer "informs" the water
through the EM coil with the digital signal for 10 min and
moves the tube back to the holding rack. It then proceeds
in the same way for the other tubes. The recordings (DTI
signal, WAT signal, NS) are "played" in a random order
determined by an Excel program on start-up of the
computer. The “informed” thrombin is then mixed with
fibrinogen and 2 aliquots of the thrombin/fibrinogen
mixture are distributed side-by-side into the wells of a
96-well microplate where coagulation starts (Fig. 1A). The
latter is read every 60 s for 1 h using a spectrophotometer
and the program saves the data to an Excel spreadsheet.
Analysis of this data was done by hand to determine
coagulation kinetics.
It was agreed at the start of the study that only pilot
phase and test phase data would be formally analyzed.
However, in the pre-pilot phase, DTI, WAT, and NS
samples each produced apparent delayed coagulation
(decreased rate of OD increase) 25% of the time,
indicating some digital effect might be occurring.
Experiments done between the pre-pilot and pilot
phases were designed to determine variance and esti-
mate a sample size sufficient to detect a 20% difference
between groups. These data were obtained with NS and
WAT treated samples only. The ABA machine pro-
duced a variance of ⬍1% standard error, suggesting
that four experiments would be sufficient to detect a
20% difference between samples. No digital effects
were seen in 17 experiments with two investigators
during this pre-pilot phase. (Fig. 2A)
During the pilot phase (October 30–November 3,
2001) Prof. Benveniste and his team conducted 16
experiments using the three-group design. While these
experiments were randomized it was possible to deter-
mine when no signal (NS) was the experimental con-
dition by observing a volt/ohm meter placed in parallel
with the signal playing from the computer to coil.
However, it was not possible to determine when the
DTI or WAT signals were being played. Coagulation
differences between WAT and NS were ⬍1%. How-
ever, in 7 of the experiments a 21–28% decrease in
mean coagulation rate was observed for DTI (Fig. 2B).
A subgroup analysis of pilot phase data showed that
all DTI effects occurred when experiments were con-
ducted by one member of Benveniste’s team (Jamal
Aissa) and that this usually occurred when using a split
sample technique in which he interrupted the opera-
tion of the ABA machine to do manual plating followed
by the automated plating. Two of the 16 experiments
done only by the ABA machine (no interruption)
showed effects when Jamal was present. In three in-
stances Jamal set up experiments and then left for the
day. None of these showed DTI effects. An example
curve of a “successful” experiment is shown in Fig. 2B.
After November 3, 2001, the independent test phase
was conducted. This consisted of 29 formal experi-
ments run by six different investigators in two different
laboratories using the same three group design con-
ducted in the pilot phase. Mean coagulation differ-
ences between groups was 1–2% and nonsignificant
(ANOVA F⫽0.04, P⫽0.96; Kruskal-Wallis Chi
square⫽0.06, P⫽0.97). The mean optical densities at
40 min were 1.08 ⫾0.1 (95% CI: 1.05–1.12), 1.08 ⫾
0.09 (95% CI: 1.05–1.11), and 1.09 ⫾0.08 (95% CI:
1.06–1.12) for DTI, water, and blank, respectively. An
example curve is seen in Fig. 2C.
After the test phase was completed, we conducted
several experiments using varying concentrations of
thrombin and fibrinogen with and without digital sig-
nals in an attempt to reproduce the curves illustrated in
25CAN SPECIFIC BIOLOGICAL SIGNALS BE DIGITIZED?
Fig. 2Bwith known chemicals. No combination of
sample splitting or concentration change exactly repro-
duced the curves of the Benveniste team. However,
adding between 100 and 200 L water or between 2
and 4 ng/mL of the thrombin inhibitor Lepirudin per
well resulted in a decreased coagulation rate (Fig. 2D).
Statistical results similar to the pilot phase experiments
can be produced with these techniques, although the
coagulation curve is flatter and more linear than those
produced in the pilot phase experiments. In addition,
the volume differences required to produce such ef-
fects are easily distinguished by visual inspection. There
was no evidence of volume or concentration irregular-
ities seen during any of the pilot phase experiments.
A detailed analysis of the digital signal conducted by
our electrical engineer found no differences between DTI
and WAT signals. A simple voltmeter (tester in Fig. 1B)
was used to measure the average signal strength being
applied to the coils. This was used to verify that a signal
was in fact being presented to the samples. The same
voltage, and thus signal intensity, was observed when each
of these signals was presented to the samples.
Verification of procedural protocols was provided by
our consultants in statistics and hematology.
APPLICATION OF THE SOCIAL MANAGEMENT
PROCESS
Ongoing evaluation by our communication and con-
flict management consultant documented the commu-
nication and feedback dynamics that operated within
the research team. Conclusions were drawn from ex-
tensive observations of interactions during both the
conduct of the experiment and research team meet-
ings, in-depth interviews with Prof. Benventiste and his
associates along with the members of the research
team. The following questions were asked and re-
sponses were recorded and later reviewed: 1) How do
you feel the study is progressing? 2) Are there particular
challenges you face in the conduct of the study? 3) How
successful/unsuccessful do you feel your (visit, etc.) has
Figure 1. A) Photograph of automated biol-analyzer (ABA).
Photograph of ABA work surface. Six Eppendorf tubes are
placed by hand, starting from far left, in each of the two
Metallic (soft iron) Racks. The robot arm (not pictured)
using Disposable tips distributes aliquots of thrombin from
the Thrombin reservoir into each of the Eppendorf tubes in
Metallic Rack 1. After filling each Eppendorf tube with a
predetermined volume of thrombin and before filling the
next Eppendorf tube in line, the robot removes the filled
Eppendorf tube from the Metallic Rack and places it in the
Signal Coil. The tube is left there for 10 min during which
time one of three signals (randomly determined) is played:
DTI, WAT, or NS (see text). At the end of 10 min the tube is
returned to the rack, and the disposable tip is ejected into the
Garbage Box. A fresh disposable tip is acquired and the next
tube in line is filled and placed in the Signal Coil. After 60
min when all tubes have been “treated” in this way, the robot
proceeds to distribute aliquots of fibrinogen from the Fibrin-
ogen reservoir into each empty Eppendorf tube in Metallic
Rack 2. After filling each Eppendorf tube with a predeter-
mined volume of fibrinogen and before filling the next tube
in line, the robot adds, from the parallel tube, a predeter-
mined volume of treated thrombin. It then mixes the two
compounds by repeated pipetting and then distributes the
resulting mixture into two side-by-side wells of the Microplate.
The robot then proceeds to the next tube down the line and
repeats the process until all six of the tubes have been
processed and aliquots from each are distributed into the
appropriate wells of the Microplate. B) Schematic diagram of
the apparatus for “informing water.” The device used for
“informing” the water consists of a standard computer fitted
with a sound card. The “line-out” outlet of the sound card is
linked to the “line-in” of a commercially available stereo hi-fi
amplifier by a stereo connection: a male minijack (3.5 mm)
plug on the sound card side and two cinch plugs on the
amplifier side. The wires from an electromagnetic coil (4
ohms) and the probes of an AC voltmeter are plugged into
the speaker outlet at the back of the amplifier. When the
prerecorded ".wav" files of either DTI or water (WAT) are
played by the computer, the signal is delivered to the coil at
4V AC for 10 min. The frequency range is that of the sound
card and the amplifier ranging from 20 Hz to 20 kHz. Not
shown here is the recording process in which the sample to be
recorded is placed between one transmitting and two receiv-
ing coils. The transmitting coil produces an electromagnetic
noise signal that is passed through the sample and received by
the receiving coils. These two receiving coils drive a differen-
tial amplifier whose output is connected to the "mic" input of
the computer sound card and the recording is achieved as is
done with normal sound through a microphone with settings
at 44.1 kHz, mono, for 3 s.
26 Vol. 20 January 2006 JONAS ET AL.The FASEB Journal
been? 4) How confident are you that a rigorous scien-
tific study can be completed? 5) Have your own per-
sonal views and beliefs been challenged? How? What
has been your response? and 6) What explanations for
the phenomena under examination do you believe to
be most plausible? Why?
Results from these interviews revealed strong agree-
ment among all interviewed that the study was progress-
ing well and certain challenges were identified and were
thoroughly discussed to everyone’s satisfaction. Prof. Ben-
veniste’s group felt their visit and set-up of the experiment
was very successful. To question #4, concerns were ex-
pressed by Prof. Benveniste’s group concerning the kinds
of interpretations one might give to findings of no signif-
icance or inconsistent findings. These concerns were
directly addressed, again to everyone’s satisfaction, in an
explicit statement by the research team regarding these
issues in a follow-up memo of October 31, 2001 (not
shown). Responses to the remaining two questions indi-
cated lack of conflict or disagreements in these areas. The
overall dynamics of the American research team’s inter-
action with the Digibio members as well as among them-
selves was respectful of differences of opinion and collab-
orative in decision-making.
A quantitative assessment of the degree of group con-
sensus within the American research team was achieved. A
21-item Consensus Scale was administered following the
conclusion of the research effort to the five members of
the American research team. Responses were obtained
from three members. This measure, which obtained reli-
ability coefficients of 0.89 in previous administrations,
assessed the degree of agreement of respondents to five
dimensions of group consensus: effectiveness of decision-
making process, individual agreement with the group
decision, group members' relations, individual effective-
ness and opportunity to participate. Results indicated a
mean scale score of 4.53/5.00 (4.61, 4.33, and 4.71),
suggesting a high degree of overall group consensus
among research team members (22). Overall ratings
among the three respondents on each of the five sub-
scales of the social consensus measure were similarly high.
Overall, all three respondents were in high agreement
with one another concerning the strong level of social
consensus obtained among the members. The social and
communication process was characterized as open in
providing opportunities for improvement in communica-
tion, establishing fair and consistent protocols for the
study, and respectful and direct resolution of differences
of opinion during the experiment. Thus, there was an
avoidance of the perils of “groupthink phenomena” (23,
24).
In addition, a set of guiding principles was estab-
lished that aided the research team in addressing some
important changes/concerns expressed by Prof. Ben-
veniste and his associates at different times during the
study as well as facilitating agreement on the interpre-
tation and judgment of the obtained findings by the
research team. Prof. Benveniste provided comments on
these test results and suggestions but no editing or
authorship on the final report for DARPA. For exam-
ple, one guideline stated, "differing interpretations of
the meaning and implications of the data are possible
and can be reflected in individual research team mem-
ber reports." This guideline permitted alternative
voices to be heard if full agreement on specific issues or
findings was not obtained.
Overall, a fair and effective communicative process
was established and maintained around research con-
ducted on a highly controversial topic. Thus, the social
management process allowed for the project to be
Figure 2. Thrombin-fibrinogen coagulation
rates optical density curves representing co-
agulation rates of the thrombin-fibrinogen
mixture exposed to various digital signals,
thrombin inhibitors and controls. E,䊐⫽
water (sample not exposed to any digital
signal); ,ƒ⫽water signal (sample ex-
posed to signal produced from plain water);
F,■⫽samples exposed to digitized signal
produced from thrombin inhibitor (DTI).
A) Example run produced by automated
biological analyzer (ABA) with no signals
running. Variation between samples was ⬍
1% over 10 runs completed on 5 different
days. B) Example run from “successful” ex-
periment conducted by Jamal Aissa. Delayed
coagulation was observed in samples ex-
posed to DTI in 11 of 23 samples analyzed,
average inhibition 24% (P⬍0.0001). C) Ex-
ample curve selected from over 40 runs
conducted by seven independent investiga-
tors on the digital effect. Variation between
samples ⬍2% throughout replication at-
tempts (P⫽0.96). D) Example coagulation
inhibition differences produced by adding 5
concentrations of the thrombin inhibitor Lepirudin (a hirudin derivative) to test samples compared with digitized water and
plain water demonstrating ability of ABA to produce a coagulation inhibition effect with a standard chemical thrombin
inhibitor. F⫽4 ng/mL; E⫽2 ng/mL; ⫽1 ng/mL; ƒ⫽400 pg/mL; ■⫽200 pg/mL; 䊐⫽zero Lepirudin present.
27CAN SPECIFIC BIOLOGICAL SIGNALS BE DIGITIZED?
completed in a fair and collegial manner in a highly
controversial area of science.
SUMMARY AND CONCLUSIONS
We found no effects from digital signals on the inhibi-
tion of thrombin/fibrinogen coagulation. We observed
apparent inhibition of thrombin/fibrinogen coagula-
tion by a digital signal when one member of the
Benveniste team conducted experiments in our labora-
tory. We did not observe systematic influences such as
pipetting differences, contamination, or violations in
blinding or randomization that would explain these
effects from the Benveniste investigator. However, our
observations do not exclude these possibilities. The
digital effects found during ABA runs were seen after
manual pipetting by the experimenter in 14 of the 16
runs. While concentration changes in fibrinogen pro-
duce statistically similar reduced coagulation rates, the
curves from such concentration changes do not mimic
those seen in the pilot experiments. In addition, vol-
ume changes were not found between samples and all
protocol data were run blind, thereby decreasing the
probability that pipetting differences occurred system-
atically between groups. Improvements in blinding (by
eliminating the NS control group) and prerandomiza-
tion control (by delivering randomization codes before
each run separate from those generated by Excel)
would not alter the basic findings that digital effects
were not reproduced by our investigators.
Prof. Benveniste died on October 3, 2004. Before he
passed away, however, he posited unknown interactions
with digital signals that produce these effects and states
that he observed similar experimenter variability in his
laboratory (personal communication). He stated that
certain individuals consistently get digital effects and
other individuals get no effects or block those effects.
While it is possible that other, unknown “experimenter”
factors, such as the influence of chemical residues, ener-
getic emanations or intentionality from individual exper-
imenters could be an explanation for these findings, we
did not test these hypotheses nor developed a framework
that would control for such factors. Without such a
framework, continued research on this approach to digi-
tal biology would be at worst an endless pursuit without
likely conclusion, or at best premature (25).
In addition, we feel that this first incorporation of a
systematic social management process into research on a
controversial issue has proven useful and is a better
skeptical approach to such claims than “armchair” criti-
cism. Future research teams exploring claims that lie
outside the normal paradigms of science should consider
incorporating such procedures into their investigations.
We are pleased to acknowledge with appreciation the late
Jacques Benveniste, and his colleagues Didier Guillonnet, and
Jamal Aissa for their time, open cooperation, patience, flexi-
bility and assistance in setting up, adjusting and conducting
the experiments. Thanks to Ronald A. Chez for his review and
editing of the paper.
1. Davenas, E., Beauvais, J., Oberbaum, M., Robinzon, B., Mia-
donna, A., Tedeschi, A., Pomeranz, B., Fortner, P., Belon, P.,
Sainte-Laudy, J., et al. (1988) Human basophil degranulation
triggered by very dilute antiserum against IgE. Nature (London)
333, 816–818
2. (1988) When to believe the unbelievable. Nature (London) 333,
787
3. Maddox, J. (1988) High-dilution" experiments a delusion. Na-
ture (London) 334, 287–289
4. Coles, P. (1988) Benveniste controversy rages on in the French
press. Nature (London) 334, 372
5. Ovelgonne, J. H., Bol, A. W., Hop, W. C., and van Wijk, R. (1992)
Mechanical agitation of very dilute antiserum against IgE has no
effect on basophil staining properties. Experientia 48, 504–508
6. Hirst, S. J., Hayes, N. A., Burridge, J., Pearce, F. L., and
Foreman, J. C. (1993) Human basophil degranulation is not
triggered by very dilute antiserum against human IgE. Nature
(London) 366, 525–527 (see comments)
7. Schiff, M. (1994) The Memory of Water, Thorsons (division of
Harper Collins Publishers), London
8. Belon, P., Cumps, J., Ennis, M., Mannaioni, P., Sainte-Laudy, J.,
Roberfroid, M., and Wiegant, F. (1999) Inhibition of human
basophil degranulation by successive histamine dilutions: Re-
sults of a European multi-centre trial. Inflamm. Res. 48, S17–S18
9. Brown, V., and Ennis, M. (2001) Flow-cytometric analysis of
basophil activation: inhibition by histamine at conventional and
homeopathic concentrations. Inflamm. Res. 50, S47–S48
10. Benveniste, J., Aissa, J., and Guillonnet, D. (1999) The molecu-
lar signal is not functional in the absence of "informed water."
FASEB J. 13, A163 (abstr.)
11. Thomas, Y., Schiff, M., Belkadi, L., Jurgens, P., Kahhak, L., and
Benveniste, J. (2000) Activation of human eurtrophils by elec-
tronically transmitted phorbol-myristate acetate. Med. Hypotheses
54, 33–39
12. Walleczek, J. (2000) Self-organized Biological Dynamics & Nonlinear
Control, Cambridge University, Cambridge
13. Hammer, M., and Rogan, R. (1997) Negotiation models in crisis
situations: The value of a communication based approach. In
Dynamic Processes of Crisis Negotiation: Theory, Research and Practice
(Rogan, R., Hammer, M., and Van Zandt, C., eds) pp. 9–23,
Praeger, Westport, CT
14. Rogan, R., and Hammer, M. (2002) Crisis/hostage negotiations:
a communication-based approach. In Law Enforcement, Commu-
nication, and Community (Giles, H., ed) pp. 229–254, John
Benjamins, Amsherdam
15. Cissna, K., and Sieburg, E. (1981) Patterns of interaction
confirmation and disconfirmation. In Rigor and Imagination:
Essays from the Legacy of Gregory Bateson (Wilder, C., and Weak-
land, J., eds) pp. 253–282, Praeger, New York
16. Ruesch, J., and Bateson, G. (1951) Communication: The Social
Matrix of Psychiatry, WW Norton, New York
17. Watzlawick, P., Beavin, J., and Jackson, D. (1967) Pragmatics of
Human Communication, WW Norton, New York
18. Hammer, M. (1996) Burzynski antineoplaston case study: con-
flict issues and recommendations. In Office of Alternative
Medicine, NIH, Bethesda, MD
19. Hammer, M., and Jonas, W. (2004) Managing social conflict in
CAM research: the case of antineoplastons. Integr. Cancer Ther. 3,
59–65
20. Hammer, M. (1996) The management of dispute and judgment
process in controversial complementary and alternative medicine
research. In Office of Alternative Medicine, NIH, Bethesda, MD
21. Hammer, M. (1997) Practice outcomes monitoring and evalua-
tion system (POMES): Report of recommendations concerning
the POMES system. In Office of Alternative Medicine, NIH,
Bethesda, MD
22. DeStephens, R., and Hirokawa, R. (1988) Small group consen-
sus: stability of group support of the decision task process, and
group relationships. Small Group Behavior 19, 227–239
23. Janis, I. (1982) Groupthink, Houghton Mifflin, Boston
24. Janis, I. (1989) Crucial Decisions: Leadership in Policymaking and
Crisis Management, The Free Press, New York
25. Stent, G. S. (1972) Prematurity and uniqueness in scientific
discovery. Sci. Am. 227, 84–93
Received for publication March 1, 2005.
Accepted for publication September 23, 2005.
28 Vol. 20 January 2006 JONAS ET AL.The FASEB Journal
