The FASEB Journal Hypothesis

Can specic biological signals be digitized?

Wayne B. Jonas,*,,1,2 John A. Ives,*,,1 Florence Rollwagen,,1 Daniel W. Denman, Kenneth Hintz, 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; 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 Defense Advanced Research Projects Agency, we attempted to replicate the data of Professor Jacques Benveniste that digital signals recorded on a computer disc produce specic biological effects. The hypothesis was that a digitized thrombin inhibitor signal would inhibit the brinogen-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 conict management. The social management approach was an adaptation of interactional communication 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, neuroscientist and conict 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 specic biological signals be digitized? FASEB J. 20, 2328 (2006)

results (3). Charge and counter-charge ensued (mostly in the press) (4) while subsequent attempts to scientifically investigate the claim produced negative (5), equivocal (6, 7), and positive (8, 9) reports. While the topic largely disappeared from the scientic literature, it continues to be discussed and evaluated by the public in prominent media places such as the BBC and ABCs 20/20 (http://www.abcnews.go.com/sections/2020/ GiveMeABreak/GMAB_homeopathy_040130-1.html). During the ensuing years Prof. Benveniste has pursued 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 specic manner (10, 11) (www.digibioSA.com). While it is known that molecular interactions in biological systems can be inuenced by low-level electromagnetic elds (12), the ability to specify and deliver signals through electromagnetic mechanisms has not otherwise been reported. The implications for digitally mimicked biological 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. Benvenistes reports.

Key Words: social management of science controversial science claims electromagnetic signals digital biology thrombin brinogen coagulation

MANAGING SOCIAL CONFLICT The controversial nature of the topic and the handling of previous research claims of Prof. Benveniste by the scientic community required, in addition to normal scientic methods of investigation, a social and communication management process that was capable of dealing with conicting interpersonal dynamics among vested parties
1 The views, opinions, and assertions expressed in this article are those of the authors and do not reect ofcial 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 Advanced 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

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 scientic process of waiting for independent replication and sent a team to investigate the claim in Benvenistes laboratory. Under emotionally charged conditions, they performed and subsequently published their
0892-6638/06/0020-0023 FASEB

23

in the research effort. The approach we used represents a modication of a communication-based process originally developed within the crisis management arena that focuses on the functional meaning of discourse taking place under conditions of disagreement and emotional intensity (13, 14). The process is an adaptation from interactional communication theory and identies four conict discourse frames that focus communication among parties around interactional triggers of conict escalation and de-escalation (1517). 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 inuence, 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).

empathize with the other, simplied thinking, and overreliance 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 conict resolution professional into the research team. His role was to observe, monitor, and evaluate the social dynamics during implementation of the research project. This involved several functions, including 1) intervention into group decision-making processes around both substantive issues and assessment of attunement, face, and emotional discord dimensions; 2) application of conict communication skills (e.g., active listening) to key prioritization 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 established), execution of the study (e.g., how are protocols actually implemented, how are changes/modications made to the previously agreed upon research protocols), and interpretation and evaluation of the research ndings (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 observations 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. Benventistes claims, requested of us by the Department of Defense.

SOCIAL CONFLICT IN MEDICAL RESEARCH We previously developed a process for the adaptation and application of this social and communication management process to controversial scientic research topics. Specically, under the direction of Dr. Wayne Jonas, then director of the Ofce of Alternative Medicine (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 conicts 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 conict 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 conict management processes to be integrated into a practice outcomes monitoring and evaluation system (POMES) designed to address research questions related to controversial cancer therapies (21). These ndings indicated that difculties negatively affecting the conduct of research efforts on controversial topics not only reect differences of opinion on substantive issues (e.g., research protocols, appropriate statistical analysis) but, more important, reect social conict among the parties. This leads to interpretive and judgment biases that result in selective ltering and recall of data, dismissal of nonconforming data, increased mistrust of the motives of the other party, reduced ability to
24 Vol. 20 January 2006

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 communication and management process to be used. The protocol was reviewed and agreed to by Prof. Benveniste and his staff, including the criterion that he would serve as a consultant and not an author of any nal 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/conict expert was recruited. Feedback and modications of the experimental proJONAS 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 Benvenistes 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 modications to the ABA program and protocol were implemented and veried at that time, including blinding and randomization procedures. These same two individuals, along with Prof. Benveniste, returned on October 30 through November 3, 2001 to run the rst pilot phase under the modied protocol. The test phase was then conducted by our research team using the same protocol, but independent of Prof. Benvenistes 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 P value set at 0.05 for hypothesis determination. According to Prof. Benveniste, a biological system that can be used to detect digital signals is the inhibition of brinogen-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 inhibitor (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 delivery 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 electromagnetic (EM) signal or water signal. After exposure, the samples are mixed with brinogen (24 mg/mL in 0.1% sodium azide, to prevent bacterial growth) and distributed 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 brinogen and thrombin, and their loading into ABA reservoirs. The basic ABA procedure is as follows. After placement of brinogen 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 brinogen and 2 aliquots of the thrombin/brinogen mixture are distributed side-by-side into the wells of a 96-well microplate where coagulation starts (Fig. 1A). The
CAN SPECIFIC BIOLOGICAL SIGNALS BE DIGITIZED?

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 estimate a sample size sufcient to detect a 20% difference between groups. These data were obtained with NS and WAT treated samples only. The ABA machine produced a variance of 1% standard error, suggesting that four experiments would be sufcient 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 determine when no signal (NS) was the experimental condition 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%. However, in 7 of the experiments a 2128% 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 conducted by one member of Benvenistes team (Jamal Aissa) and that this usually occurred when using a split sample technique in which he interrupted the operation 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 instances 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 experiments run by six different investigators in two different laboratories using the same three group design conducted in the pilot phase. Mean coagulation differences between groups was 12% and nonsignicant (ANOVA F0.04, P0.96; Kruskal-Wallis Chi square0.06, P0.97). The mean optical densities at 40 min were 1.08 0.1 (95% CI: 1.051.12), 1.08 0.09 (95% CI: 1.051.11), 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 brinogen with and without digital signals in an attempt to reproduce the curves illustrated in
25

Fig. 2B with known chemicals. No combination of sample splitting or concentration change exactly reproduced 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 atter and more linear than those produced in the pilot phase experiments. In addition, the volume differences required to produce such effects are easily distinguished by visual inspection. There was no evidence of volume or concentration irregularities 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. Verication of procedural protocols was provided by our consultants in statistics and hematology.

APPLICATION OF THE SOCIAL MANAGEMENT PROCESS Ongoing evaluation by our communication and conict management consultant documented the communication and feedback dynamics that operated within the research team. Conclusions were drawn from extensive observations of interactions during both the conduct of the experiment and research team meetings, in-depth interviews with Prof. Benventiste and his associates along with the members of the research team. The following questions were asked and responses 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 lling each Eppendorf tube with a predetermined volume of thrombin and before lling the next Eppendorf tube in line, the robot removes the lled 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 lled 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 brinogen from the Fibrinogen reservoir into each empty Eppendorf tube in Metallic Rack 2. After lling each Eppendorf tube with a predetermined volume of brinogen and before lling the next tube in line, the robot adds, from the parallel tube, a predetermined 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 tted with a sound card. The line-out outlet of the sound card is linked to the line-in of a commercially available stereo hi-
26 Vol. 20 January 2006

amplier by a stereo connection: a male minijack (3.5 mm) plug on the sound card side and two cinch plugs on the amplier 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 amplier. When the prerecorded ".wav" les 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 amplier 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 receiving 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 differential amplier 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.
JONAS ET AL.

The FASEB Journal

Figure 2. Thrombin-brinogen coagulation rates optical density curves representing coagulation rates of the thrombin-brinogen mixture exposed to various digital signals, thrombin inhibitors and controls. E, water (sample not exposed to any digital signal); , water signal (sample exposed 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 experiment conducted by Jamal Aissa. Delayed coagulation was observed in samples exposed to DTI in 11 of 23 samples analyzed, average inhibition 24% (P0.0001). C) Example curve selected from over 40 runs conducted by seven independent investigators on the digital effect. Variation between samples 2% throughout replication attempts (P0.96). 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; 400 pg/mL; 200 pg/mL; zero Lepirudin present.

been? 4) How condent are you that a rigorous scientic study can be completed? 5) Have your own personal 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 agreement among all interviewed that the study was progressing well and certain challenges were identied and were thoroughly discussed to everyones satisfaction. Prof. Benvenistes group felt their visit and set-up of the experiment was very successful. To question #4, concerns were expressed by Prof. Benvenistes group concerning the kinds of interpretations one might give to ndings of no significance or inconsistent ndings. These concerns were directly addressed, again to everyones 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 indicated lack of conict or disagreements in these areas. The overall dynamics of the American research teams interaction with the Digibio members as well as among themselves was respectful of differences of opinion and collaborative in decision-making. A quantitative assessment of the degree of group consensus within the American research team was achieved. A 21-item Consensus Scale was administered following the conclusion of the research effort to the ve members of the American research team. Responses were obtained from three members. This measure, which obtained reliability coefcients of 0.89 in previous administrations, assessed the degree of agreement of respondents to ve dimensions of group consensus: effectiveness of decisionmaking process, individual agreement with the group decision, group members' relations, individual effectiveCAN SPECIFIC BIOLOGICAL SIGNALS BE DIGITIZED?

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 ve subscales 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 communication, 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 established that aided the research team in addressing some important changes/concerns expressed by Prof. Benveniste and his associates at different times during the study as well as facilitating agreement on the interpretation and judgment of the obtained ndings by the research team. Prof. Benveniste provided comments on these test results and suggestions but no editing or authorship on the nal report for DARPA. For example, one guideline stated, "differing interpretations of the meaning and implications of the data are possible and can be reected in individual research team member reports." This guideline permitted alternative voices to be heard if full agreement on specic issues or ndings was not obtained. Overall, a fair and effective communicative process was established and maintained around research conducted on a highly controversial topic. Thus, the social management process allowed for the project to be
27

completed in a fair and collegial manner in a highly controversial area of science.

1.

SUMMARY AND CONCLUSIONS We found no effects from digital signals on the inhibition of thrombin/brinogen coagulation. We observed apparent inhibition of thrombin/brinogen coagulation by a digital signal when one member of the Benveniste team conducted experiments in our laboratory. We did not observe systematic inuences 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 brinogen produce statistically similar reduced coagulation rates, the curves from such concentration changes do not mimic those seen in the pilot experiments. In addition, volume changes were not found between samples and all protocol data were run blind, thereby decreasing the probability that pipetting differences occurred systematically between groups. Improvements in blinding (by eliminating the NS control group) and prerandomization control (by delivering randomization codes before each run separate from those generated by Excel) would not alter the basic ndings 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 inuence of chemical residues, energetic emanations or intentionality from individual experimenters could be an explanation for these ndings, 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 digital biology would be at worst an endless pursuit without likely conclusion, or at best premature (25). In addition, we feel that this rst 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 criticism. 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, exibility and assistance in setting up, adjusting and conducting the experiments. Thanks to Ronald A. Chez for his review and editing of the paper.

2. 3. 4. 5. 6.

7. 8.

9. 10. 11.

12. 13.

14.

15.

16. 17. 18. 19. 20. 21.

22. 23. 24. 25.

Davenas, E., Beauvais, J., Oberbaum, M., Robinzon, B., Miadonna, 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 (1988) When to believe the unbelievable. Nature (London) 333, 787 Maddox, J. (1988) High-dilution" experiments a delusion. Nature (London) 334, 287289 Coles, P. (1988) Benveniste controversy rages on in the French press. Nature (London) 334, 372 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 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, 525527 (see comments) Schiff, M. (1994) The Memory of Water, Thorsons (division of Harper Collins Publishers), London 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: Results of a European multi-centre trial. Inamm. Res. 48, S17S18 Brown, V., and Ennis, M. (2001) Flow-cytometric analysis of basophil activation: inhibition by histamine at conventional and homeopathic concentrations. Inamm. Res. 50, S47S48 Benveniste, J., Aissa, J., and Guillonnet, D. (1999) The molecular signal is not functional in the absence of "informed water." FASEB J. 13, A163 (abstr.) Thomas, Y., Schiff, M., Belkadi, L., Jurgens, P., Kahhak, L., and Benveniste, J. (2000) Activation of human eurtrophils by electronically transmitted phorbol-myristate acetate. Med. Hypotheses 54, 3339 Walleczek, J. (2000) Self-organized Biological Dynamics & Nonlinear Control, Cambridge University, Cambridge 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 Rogan, R., and Hammer, M. (2002) Crisis/hostage negotiations: a communication-based approach. In Law Enforcement, Communication, and Community (Giles, H., ed) pp. 229 254, John Benjamins, Amsherdam Cissna, K., and Sieburg, E. (1981) Patterns of interaction conrmation and disconrmation. In Rigor and Imagination: Essays from the Legacy of Gregory Bateson (Wilder, C., and Weakland, J., eds) pp. 253282, Praeger, New York Ruesch, J., and Bateson, G. (1951) Communication: The Social Matrix of Psychiatry, WW Norton, New York Watzlawick, P., Beavin, J., and Jackson, D. (1967) Pragmatics of Human Communication, WW Norton, New York Hammer, M. (1996) Burzynski antineoplaston case study: conict issues and recommendations. In Ofce of Alternative Medicine, NIH, Bethesda, MD Hammer, M., and Jonas, W. (2004) Managing social conict in CAM research: the case of antineoplastons. Integr. Cancer Ther. 3, 59 65 Hammer, M. (1996) The management of dispute and judgment process in controversial complementary and alternative medicine research. In Ofce of Alternative Medicine, NIH, Bethesda, MD Hammer, M. (1997) Practice outcomes monitoring and evaluation system (POMES): Report of recommendations concerning the POMES system. In Ofce of Alternative Medicine, NIH, Bethesda, MD DeStephens, R., and Hirokawa, R. (1988) Small group consensus: stability of group support of the decision task process, and group relationships. Small Group Behavior 19, 227239 Janis, I. (1982) Groupthink, Houghton Mifin, Boston Janis, I. (1989) Crucial Decisions: Leadership in Policymaking and Crisis Management, The Free Press, New York Stent, G. S. (1972) Prematurity and uniqueness in scientic discovery. Sci. Am. 227, 84 93 Received for publication March 1, 2005. Accepted for publication September 23, 2005.

28

Vol. 20

January 2006

The FASEB Journal

JONAS ET AL.