Nearly every major new technology appears to be part of an established continuum of experience when in fact it isn't; instead, it opens the gate on a vista of new ignorance that offers unanticipated risks. Genetically altered plants offer a new vista; they are in a different framework from past experience and present a fresh risk that we will unwittingly take in our rush to embrace technical virtuosity. Gene splicing is unrelated to established plant breeding, hybrids, and grafted plants, just as the horse-drawn carriage was unrelated to the automobile. While we apply the term horsepower to both horse-drawn carriages and cars, the technological similarity is insignificant. The same misuse of metaphors applies to gene technology; it is not a linear offspring of breeding.

It is not my intention to use fear to intimidate the reader on the subject of genetic food alteration. I wish to draw a reasonable line so that we know where that line lies and we can consciously choose whether to cross it. As you will learn from further reading, our federal government has already decided to erase that line, under pressure from former Vice President Dan Quayle's Committee on Competitiveness.

In 1995, a man carried radioactive cesium in his pocket, in an unprotected package, from East Europe to Germany, where he had it tested. It was apparently sold to him by a secret police agent. The courier died promptly of exposure to radioactivity. He was reportedly ignorant of the lethal contents of his package. This is one kind of ignorance: ignorance of a relevant fact. People near him on the train or plane had the same kind of ignorance.

Ignorance of relevant fact is the first order of ignorance. It belongs to a vast domain and causes the death of thousands of Americans annually and maims even more. Many people have died because they didn't know that their foot brake was defective, that the contents of the bottle they were pouring were poison, that the hustler on the street had a loaded gun, or that the driver of the car coming at them couldn't see them.

If the cesium carrier had known that he was carrying radioactive cesium but didn't understand that a significant dosage of radioactivity is dangerous, this would be an example of second-order ignorance. In this case, the relevant fact is known, but it is not understood.

Third-order ignorance is different from not having the relevant fact and not understanding it, third-order ignorance involves both the impossibility of knowing the relevant fact and the impossibility of understanding the significance of its occurrence. I will show that genetically altered foods are in the domain of the third order.

There is a large body of harmful ignorance that falls into the second order. Among the American deaths and hospitalizations from this source come some of the common tragedies where people don't understand the relationship between vehicle speed and braking distance, or the decline in visual acuity that comes from aging, or the transmission mechanism of the AIDS virus, or the relationship between guns in the household and suicide and murder (the mere presence of guns accounts for more than half of all American suicides).

There are many cases of maiming and death that appear to be second-order ignorance, but aren't. Most people who smoke and drink are not ignorant of the consequences; instead, they are making a choice. These actions should not be confused with second-order ignorance.

Many deaths and illnesses in America result from a second-order ignorance resembling the plight of the cesium traveler. Examples include farm workers who handle pesticides, roofers who work without masks, people who use electricity near ground water, and young people who volunteer for frontline duty in the military (they believe they are invincible).

So far, we have relied on common sense associated with personal experience to argue that ignorance with harmful consequences can be divided into several meaningful categories. The purpose of this multi-step progression is to go from the realm in which each of us has had the experience of harm, such as slipping on a sidewalk or getting an electric shock, to the third order of ignorance, which has little to do with personal experience. This order is an important and meaningful category that requires thinking rather than common sense and experience.

Third-order ignorance is a state in which the relevant fact that causes harm is not accessible and the understanding of the relevant system is not possible. The only source of evidence for this argument is to examine history. I have to show that there were points in history where third-order ignorance existed and people acted in spite of it, causing harm.

If examples from history are irrelevant to the reader or of minor importance, then read no further. If the reader believes that history offers meaningful examples that are relevant to everyday life, then consider the three examples I have chosen: thalidomide, radium and DDT. I will ask of you that you generalize from these examples to the broader recognition that third-order ignorance is more common in the introduction of new technology than just these three cases.

Thalidomide is a drug that was first clinically tested in the late 1950s as an anti-depressant and anti-nausea aid. Few harmful side effects were found through normal mice testing. After about three years of testing on broad human populations, mostly in West Germany, doctors began giving it to women for morning sickness. It had positive effects on the morning sickness, and most doctors were not worried about possible effects on a fetus because of a prevailing belief that the placenta was an effective barrier against nearly everything in the mother's blood.

About 8,000 pregnant women are known to have taken the drug before it was recognized, in 1962, that thousands of babies were being born without limbs, eyes, or closed stomachs as a result of their mother's ingestion of thalidomide during pregnancy.

Thorough research, after the fact, determined that many drugs pass from the mother's bloodstream through the placenta, and that in the case of thalidomide there was an additional problem unrelated to the dosage that the fetus was receiving. The problem was chirality, a matter of chemistry discovered a century earlier by Louis Pasteur.

Many organic molecules (meaning nearly everything in the human body) are either right-handed or left-handed in their physical structure, and this has a lot to do with the way they interact with other molecules. Nearly all are actually left-handed. Thalidomide was a synthesized chemical that had an equal number of left- and right-handed molecules. The right-handed ones were destructive in the developing cells of the fetus.

Two new biological systems were recognized: transport of chemicals through the placental barrier and the effect of chiral molecules on fetal growth.

Third-order ignorance. Ignorance of these facts and their meaning.

Marie Curie and her husband first detected and identified radium in 1898. Marie coined the word radioactive at the time. By 1902, they had concentrated one-tenth of a gram of it from several tons of pitchblend (which also includes uranium). By 1910, Mme. Curie had enough information to know that it was a white-powdered metal. It is the decayed by-product of thorium, uranium, and actinium. The radioactive decay of radium is from alpha and gamma rays.

Radium was used on commercial watch dials, in many military devices for night measurements, and for imaging, much like an X-ray machine. It was sold to the public in medicines, bath additives, and popular healing drinks. One rich man who bought large quantities of radium elixir in the late 1920s found that it increased his sexual virility.

In an American show of support for Mme. Curie, President Harding presented her with a full gram of radium at a reception in 1921. It came from the proceeds of a women's sufferage fund-raising effort. A lovely photograph of this lethal radioactive transaction exists.

Many physicists were cautious in their handling of radioactivity, but most ordinary citizens viewed the luminescence associated with radium as similar to the luminescence in fireflies.

Marie Curie died of leukemia in 1934. The wealthy fan of radium elixir died in 1930 from muscle and bone decay. The mechanisms by which radioactivity affects and destroys the human body are still poorly understood. Beginning on August 6, 1945, a large sample of humans (tens of thousands at one time) were exposed to high levels of radiation in Hiroshima and Nagasaki. Since that time, scientists have been able to develop measurements of the lethal and sub-lethal dosages of human exposure to radiation. The mechanisms by which radium harms humans are still poorly understood. Radiation itself is not perceived by our senses of sight, smell or feel, but the levels of harmful exposure are increasingly recognized and appreciated.

Again, third-order ignorance. The relevant facts were unknowable at the time and the consequence inconceivable.

In the early days of industrial chemistry, specifically in 1874, chlorobenzene and chloral hydrate were brought together in the presence of sulfuric acid to form DDT. By 1939, Paul Herman Muller, a Swiss, had discovered that it damaged the nervous system of many insects that are unpopular with humans (lice, fleas, and mosquitoes).

After the war, DDT rapidly became the insecticide of choice in medical, agricultural, and every other possible setting, including household sprays. Muller received the Nobel Prize, and the agricultural millennium appeared to be at hand. Farmers had long used pesticides, such as nicotine in solution, turpentine, and pyrethrum. DDT seemed to be a new stronger version of the old standards, albeit synthesized.

There were many small, unpublicized research projects on the harmful side effects of DDT, but only the literary skills of a single 53-year-old employee of the Fish and Wildlife Department (Rachel Carson, who published Silent Spring in 1962) called attention to the fact that brown pelicans who ate fish from inland waters were unable to reproduce because their reproductive system (eggshell formation) was damaged by high concentrations of DDT in their diet. Ten years later, DDT use and production were banned.

We still have no idea of the number of biological systems affected by the long-lasting DDT molecules nor their diverse consequences. DDT is found in antarctic mammals, deep-sea fauna, and women's breasts.

Third-order ignorance. Relevant facts unknown and consequences unknowable. DDT was enthusiastically accepted.

Calgene, a company located in Davis, California, obtained a patent license to develop a tomato from a germ line that has been altered by the insertion of new DNA. Calgene raised these tomatoes and sold them in Illinois and California. The new tomato was called the Flavr Savr. It was the first gene-altered plant food on the market.

Calgene gives two reasons for focusing on tomatoes: The public dislikes fresh supermarket tomatoes ,and most people in the industry know it. Most commercial tomatoes are picked green so they can be easily shipped to market where they are sprayed with ethylene, which turns them red but does not ripen them.

Second, tomatoes are the most commonly eaten vegetable in America. Most adults eat the equivalent of one every other day. Only one of these tomatoes per week is fresh; the other two are in catsup, pizza, salsa, juice, Bloody Marys or tomato paste.

The tasteless tomato is a recent hybrid innovation. Flavorful tomatoes were developed by the Incas from a plant related to the deadly nightshade (the leaves of tomato plants are still poisonous). The tomato is the fruit of that plant, which was brought to Europe a few decades after the Spaniards found South America. Italians eat the most tomatoes per capita.

The U.S. Government, as represented by the Department of Agriculture and the Food and Drug Administration, gave Calgene Fresh, Inc. (an Illinois subsidiary of Calgene) permission to sell Flavr Savr tomatoes to the public without labelling them to indicate their unusual technical origins. The same later was true for soy beans, corn and russet potatoes.

Under pressure from former Vice President Dan Quayle's Committee on Competitiveness, the government acted quickly to make sure that the U.S. dominates the bio-engineering field that is expected to emerge as the high technology frontier of the next century. The Depaartment. of Agriculture and the FDA were convinced that gene splicing of tomatoes was no different from breeding or hybridizing tomatoes. Numerous public statements to this effect have been made over the past decade.

Breeding and hybridizing are different from gene splicing. The former two were used by the Incas (or their predecessors) to create the original tomato. These methods are fairly ancient; certainly they are several millennia old. The breeding process is familiar to almost everyone. The field of tomatoes is screened for the desirable qualities, the plants with those qualities are mated (tomatoes have sexual propagation); and in the few cases where a desirable plant survives, the process is repeated. When a desirable plant is developed that is incapable of reproduction, the original seed combination can be maintained and repeated for new crops. This is called hybridization, and it gives us mules. Both breeding and hybridization can create toxic and undesirable plants.

Gene splicing is nothing like either of these. Calgene technicians removed the DNA from tomato seeds in 10 to 50 micro-liter units and mixed into it one or more proteins called restriction enzymes. Each DNA strand is a very long molecule that curls up into a tight clump. The restriction enzyme cuts the DNA at two specific points, which removes a gene, and the DNA strand is then reattached with a polymerase enzyme. It is then replaced in the seed (by being fired at high speed into the cell nucleus) and grown. In some gene splicing, a short gene code is inserted in a gene that has the effect of turning it off. In some cases, a new DNA sequence is delivered to the cell nucleus by a virus.

The role of many individual genes is to turn on and off protein production in the development of the plant or animal. In the few highly simplified plant and animal cases where the actions of many of the genes are known, the planaeria (a worm) for instance, it turns out that two-thirds of the genes don't do anything recognizable. They seem to be evolutionary remnants that are unexplainable. The remaining genes activate a process, such as growing skin cells, at a specific point in the cell division and then turn it off at a later point. Active genes generally produce proteins that serve multiple functions, and most times they work in conjunction with other genes.

In breeding and hybridization, it is not known what genes are added, subtracted, or changed; only the final plant product is known. In gene splicing, the gene that is removed, altered, or deactivated is usually known to play a specific role. In the Flavr Savr tomato strain. the gene that is deactivated is one that turns on the decay process in the tomato pulp; this is the process that ultimately begins the new seeds' growth.

One problem with gene splicing is that we don't know much about the way genes interact with each other. It is an unknown domain. We might not have a comprehensive answer for half a century. We do know, however, that there are several specific ways that gene splicing is clearly different from breeding and hybridizing:

First, we have more than two thousand years of human experience to rely on in breeding. If there had been some kind of obvious harm as a result, we would have seen it by now. Breeders sense the clues for danger to look out for in new breeds that survive to full grown status. The experience of thousands of years of breeding does not carry over to gene splicing.

Second, even if Flavr Savrs didn't turn out to be harmful, future splicing that uses genes from other species (such as another commercial tomato that is planned to include fish genes) may be harmful.

Third, breeding and hybridization have inherent limitations because the cell nucleus remains intact; gene splicing transcends this clearcut limitation with unknown effects.

I can imagine a few scenarios that would result in toxic consequences. Genes that normally don't turn on do so in a few plants when the growing weather is unusual, when an unusual bacteria or pest attaches to the plant, or when an unusual soil condition is present. In the absence of the Flavr Savr&endash;cut gene, the toxic gene remains turned on and solanine (a poison common to the nightshade family) accumulates in the tomato seeds.

While the tomato is probably the first gene-altered food product to reach the market, other consumable products of gene technology have entered the market in the past, including one with a toxic outcome.

About one-third of the rennet in American cheese is produced from vats of gene-altered bacteria. No known harmful effects have been detected or publicized. On the other hand, a popular nutritional supplement (not actually a vitamin) called L-Tryptophan, an amino acid, has long been consumed for insomnia. In the late 1980s, a Japanese animal-supplement company, Showa Denko, began producing and marketing L-Tryptophan made in vats from gene-altered bacteria licensed to them by a biologist at M.I.T. According to reports by the Center for Disease Control, 23 Americans died from consuming Showa Denko's L-Tryptophan, and over 2,000 were injured permanently. The syndrome in those injured consisted of scederma (a disintegration of the connective tissue), hair loss, constant muscle and joint pain, and peripheral neuropathy (numbness in the hands and feet).

In the first trial on this matter in San Diego in early 1993, expert witnesses were completely unable to identify the problem with the Showa Denko production. A single batch was found to have caused the tragic human outcomes. Samples of that toxic batch were carefully set aside at the time of production and stored. Parts of these samples have been tested extensively with no meaningful results. Rats have been fed the toxic amino acid with no harmful effects. The first trial resulted in over $1 million for the plaintiff and settlements since that time have been larger. The total settlement exceeded $2 billion.

The United States was the first country to allow its citizens to consume gene-altered food without labelling. The rapidity and success of this commerical introduction is unknown. The Calgene tomato failed in the market and so may some of its successors, delaying the issue for a few years. Nevertheless, the problem is at hand and we are confronted with it.

Since this is a new technology that takes us into a domain of third-order ignorance, we should each consciously decide the level of risk we are willing to take with our lives and our health. First, we should ask that gene-spliced foods be labeled all the way to the final user, so each of us can decide what risks we are willing to take with a new technology. Unfortunately, the FDA in October 1993 ruled that labeling is unnecessary unless the new food contains DNA from known allergens. Only Congress or a national health disaster will change this decision.

Second, many of us would like to see more extensive, serious testing of new foods. The Gnaizda (a friend of mine) testing approach would be to have all the board members and senior staff of Calgene eat five Flavr Savr tomatoes a day for one year. The Phillips test would be to have two sets of three thousand pigs, each of which would eat five tomatoes a day for a year, but only one group of which would eat Flavr Savrs. This amount of time is sufficient to allow the birth of offspring.

Third, following the recommendation of Henry I. Miller in the December 3, 1993, Science magazine, we should undertake national research to deliberately explore the worst-case scenarios of gene-altered cells, particularly in regard to allergies. Dr. Miller's article in Science was reasonable, but he is the main architect in the FDA of the hands-off policy and an on-going champion of unbridled genetic engineering.

In practical terms, if we want to avoid these untested foods, we are confined to buying food from natural foods stores that make an effort to identify their growers and to reading articles that purport to trace the introduction of gene-altered foods. "Organic" has been defined by the Department of Agriculture to exclude gene altered food products.

This current technological situation is similar to the three earlier ones already discussed. Radium luminescence was considered to be a harmless version of firefly light; Thalidomide was just an ordinary sedative, safe for pregnant women because of the impermeable placental barrier; and DDT was just a better pesticide than turpentine. Gene-splicing, we are told by our government, is just like breeding.

Propelled by a fear that Japan and Germany might get an economic lead in a new technology, and convinced, as we have been many times before, that a new technology is not really different from the old ones, we are rushing to enter a third-order domain of ignorance that almost certainly will harm us.