W ho is going to pay the externalized costs of GMOs?
             E. Ann Clark, Plant Agriculture, University of Guelph(14), Guelph, ON (eaclark@uoguelph.ca)©2000 E. Ann Clark

At issue is the degree to which commercialization of GMOs will impose unacknowledged or at least unattributed costs, involuntarily, on society and the environment. Consumers and others should be apprised of such costs - for GMOs or any other technology - to inform their decisions on acceptance or rejection of the technology.

What are "externalities"?

What are the "externalized costs" or "externalities" of GMOs in agriculture? A standard crop budget might encompass such input costs as land, diesel, seed, and fertilizer. However, it would not include costs for off-farm impacts of agricultural practice, such as:

Off-farm impacts are "externalities" which until recently, have not been factored into the calculation of agricultural cost:benefit ratios. Only a full cost accounting can give those who are going to have to pay the bill for these externalities - society and the environment - a more complete picture of the costs as well as the benefits of GM or any other technology.

Society has already begun the process of assessing externalized costs against other industries, such as forcing automakers to install catalytic converters to reduce air pollution or obliging polluting industries to install scrubbers and other technologies to achieve set standards. The precedent already exists for obliging farmers to accept responsibility for some off-farm impacts, as through manure management zones in the Netherlands, pesticide management zones in California, and restrictions on prophylactic use of antibiotics in livestock feed. It is not, therefore, unreasonable to contemplate holding the purveyors of GMOs similarly accountable(3).

The economic impact of externalities

Benefit:cost analyses for a new technology - including GM - typically look very favorable at the outset, because they do not factor in externalities. That usually comes later, based on experience, and even then, the costs are typically externalized and borne involuntarily - such as by the people of Walkerton - rather than costed against the purveyors of the technology itself.

To give an idea of how dramatically externalized costs can change the apparent profitability of technology, consider work reported by Pearce (1999) on an effort to calculate the true costs of production agriculture in the UK. When currently externalized costs were included in the sum, Jules Pretty and colleagues conservatively estimated that the true cost of UK agriculture - £2.3 billion per year or £208 ha -1 - was roughly equal to the value of the food actually produced by UK agriculture, in other words, a 1:1 ratio of benefits and costs. An older study by Pimentel et al. (1992) focused on calculating the true benefit:cost ratio of biocide use in US agriculture. When they factored in currently externalized costs(4), the benefit:cost ratio dropped from a respectable 4:1 based on direct costs only, to 1.3:1, just above breakeven. Current understanding of the deleterious effects of pesticides on human health (Colborn et al., 1996; Garry et al., 1996; Porter et al., 1999) would reduce this still further. But perhaps more to the point, roughly two thirds of the externalized costs of biocide use were exacted from society at large - involuntarily. Those paying the bill for biocide use - in the form of hospitalization for accidental poisonings, for example - were not those benefitting - the purveyors of the technology. Externalized costs were imposed on society and the environment involuntarily, without their knowledge or consent.

Both of the above studies are retrospective in nature, benefitting from decades of commercial experience to derive their conclusions.

Surely, this is precisely what the Royal Commission on Genetic Engineering is doing now for New Zealand. Clearly, a pro-active and cautionary approach to technology is the intent of the precautionary principle, which virtually all nations of the world have accepted and acknowledged by international agreement. Accordingly, let us start from the premise that scientists - or at least those who are publicly funded - have a responsibility to objectively and rigorously assess the risks as well as the benefits of agricultural genetic engineering, and to communicate their findings for the benefit of the society that pays us.

StarLink corn: a case study in externalized costs

StarLink corn has become a "mega problem"at the highest levels in the GM sector. StarLink, an Aventis (France) product, is the only GM corn hybrid which is approved for livestock feed but not for human consumption. This is because the particular kind of crystal protein in StarLink (Cry9C) is different from that in all other GM corn hybrids (Cry1Ab, Cry1Ac) and foliar Bt sprays(5). As such, the potential for allergenicity cannot be excluded on the premise of "history of safe use". This is not to say that StarLink actually will cause allergenicity - just that there is no basis for claiming it won't. In the absence of any meaningful way of actually testing for allergenicity, the USEPA is not yet prepared to authorize its use for human consumption.

Just recently, StarLink was found in Taco Bell tortillas and then in other products destined for direct human consumption(6). John Wichtrich, VP Aventis, indicated that 12% of the crop or 9 million bushels of StarLink have already been harvested and sold, apparently into the human food stream (Kaufman, 2000).

The inherent and irreconcilable incompatibility of a) growing identity-preserved grains - and most especially cross-pollinating grains, b) for large-scale commercial production/processing is starkly revealed in the current Aventis debacle. The ramifying complications radiating outwards from StarLink are illustrative of what the future holds for nations that grow GM crops.

The authorization for the commercialization of StarLink was premised on the unlikely assumption that it could and would be kept separate from all other grain(7). It was supposed to be grown by producers who agreed to keep it separate. Indeed, they were to sign an agreement to that effect. But what actually happened?

Thus, the potential for encroachment of StarLink germplasm into the human food chain existed (and indeed, still exists)(8) at every level, from seed purchase to processing and retail.

Is cross-contamination a problem unique to StarLink? The reported occurrence of triply-resistant GM canola (MacArthur, 2000)(9) in western Canada provides unambiguous evidence that the risk of genetic pollution is the same for other GM (and non-GM) crops. Indeed, pollen has always moved - up to 8 km in the case of canola - and will always move; that is the crux of the problem with gene "containment", and hence, the motivation for the "terminator" disabling technology. Prior to StarLink, genetic pollution incurred agronomic as well as marketing costs(10) which affected only farmers, because commercialized GM traits are all agronomic. Nothing currently on the market has any benefit for consumers, per se. However, should the next wave of promised GM events - functional foods - actually come to pass, the problems now occurring with StarLink will be endemic.

When the entirely predictable StarLink contamination was discovered in fall 2000, the first year it was grown commercially, massive recall problems immediately arose due to the tightly integrated nature of the processing/marketing sector in the US. Examples of the externalized costs of just this single incident, and the implications for the future, are shown in Table 1.

Lessons from StarLink

The Aventis' StarLink story has afforded us a rare opportunity to look into the future and glimpse the kinds of externalities that will accrue if GM technology advances to the so-called "next wave" of functional foods. So long as GM is limited to herbicide tolerance and pesticidal plants, genetic pollution and post-harvest mixing penalize only farmers - and specifically, those whose neighbors have opted to grow GM crops. But when the traits of interest are not agronomic but nutritional or pharmaceutical, then the impacts of cross-contamination will translate directly through to the consumers, because all costs of accomodating this technology at every stage will be added to the value of the food. Far from keeping food cheap, GM technology shows every likelihood of raising the price of food as well as contaminating it.

Table 1. Externalized costs of GM, current and potential, imposed on individual food sectors 
Description of Externalized
Future Implications
Farmers On-farm segregation, including isolation/buffers from same-crop fields and fastidious cleaning of harvest/storage/transport equipment, are costs that farmers must bear in order to separate StarLink from other types of corn. For growers of GM crops, these efforts are required for StarLink only. All other GM corn, bearing production traits only, can cross-pollinate at will without influencing the marketability of the harvested GM corn (or any other GM crop, for that matter).
  • Every out-crossing GM crop (e.g. corn; canola; cotton; potato) will have to be isolated from all other same-crop fields. Lesser value barrier crops(11) and fallow will occupy the intervening land, necessitating higher prices to make GM grain competitive. 
  • The economics of growing GM functional food grains will be further challenged by the potential for contamination, which may not simply violate the terms of the contract but could invalidate the grain for livestock or human use. 
  • Purchase of specialized cleaning systems for on-farm harvest/transport equipment will become mandatory, for use between every distinct GM field on every farm. 
  • Certified GM harvesters/carriers, with GM-based combine-mounted sensors, will proliferate, analogous to certified pesticide applicators, to guarantee safety from at-harvest and post-harvest contamination. 
  • A rigorous paper trail for each GM-sown field, coupled with unannounced audit inspections by the processors/retailers, will become mandatory. 
  • With consolidation in the elevator sector, identifying elevators able to handle specific GM crops, within a realistic hauling distance, will be a major determinant of which GM grain type(s) are grown.
  • Grain Elevators StarLink corn was received by about 260 elevators, of which half reportedly forwarded the grain for direct human use (Kaufman, 2000). Elevators must now absorb the cost of genetic testing - load-by-load - before allowing grain to enter the elevator. Testing each load lengthens the wait to unload and backs up the harvest process - at the expense of the growers. Rejected loads must be identified and kept separate or marketed elsewhere. 
  • Every load of incoming grain will be routinely tested to maintain the obligate paper trail of identity preservation. 
  • Separate facilities will be established for those types of GM grain (e.g. specific pharmaceuticals, nutritionals, plastics, vaccines, etc.) which are not delivered directly to the processor. 
  • Transport will involve dedicated trucks/rail cars to minimize risk of contamination. 
  • The requirement for grain delivery directly to the processor(s) will limit the geographic region where specific GM grains can be grown, and will add costs for specialized hauling. 
  • Infrastructural costs to accomodate these changes will be added to the value of the grain. 
  • Processing Industry Because storage and processing are so heavily concentrated in a few central facilities, a variety of different products could have been contaminated by corn pollinated by StarLink. In response, several potentially contaminated processed foodstuffs were recalled, and the possibilty remains for further recalls.
  • Every load of arriving grain will be routinely tested, not simply for dust or mold or rocks, but for genetic purity. Genetic cross-contamination could involve not just StarLink, but any of the various functional food-trait cultivars. 
  • With GM contamination coming from every direction, in the field and during post-harvest storage and transport, it will be difficult-to-impossible to assure purity. Industry will lobby vigorously for relaxed thresholds, including an urgent pitch for terminator-style disabling technology, on the grounds that pollen containment is impossible. 
  • Costs of testing, including rejected loads, will be added to the products. 
  • Off-shore (non-GM) sources of grain/produce/meat/milk will become increasingly valuable.
  • Government  Monitoring of compliance is required to ensure identity preservation of the segregated grain. Relying on the seed trade to elicit compliance from their grower clients clearly hasn't worked (e.g. StarLink). Extensive government involvement, both in monitoring and in ensuring segregation of unapproved grain, was acknowledged by Glickman, Secretary of Agriculture (Science News, 25 Sept 2000)
  • Continual monitoring of the genetic purity of GM grains will become obligatory. 
  • Computerized tracking systems and intrusive audit-trails will become mandatory at every step of the process. 
  • As evidence of StarLink-like errors mounts, possibly incurring actual human harm, government will be obliged to assume responsibility for monitoring production, sale, and movement of the hundreds of genetically distinct functional food-type cultivars and hybrids they have allowed to enter the marketplace - all the way down to the individual farm and field level. 
  • Costs of implementing this monumental tracking effort will be passed on to taxpayers and consumers.
  • Retail Stores or Fast Food Outlets.  Retail chains and 7000 Taco Bell outlets absorbed the costs of emptying shelves ( 2.5 million boxes of just taco shells alone; Pollack, 2000), and then packing up and shipping out the offending products, hopefully without alarming their customers.
  • Demand for genetic ID testing will mushroom, as markets, restaurants, and fast food outlets compete to assure their customers of the genetic purity of their products. 
  • Genetic purity labels will become as commonplace as nutritional or country-of-origin labels today 
  • Costs of so-doing will be added on to the bill 
  • Taxpayers Consumers The US government, together with Aventis, has agreed to purchase all remaining StarLink grain, including contaminated adjoining fields, from the 2000 season - at a premium no less - at a purported cost of $100 million. Although Aventis claims to be paying for the buyout, US taxpayers will presumably be paying for all other costs related to detailed monitoring, accessing, identifying, acquiring, and disposing of the 2000 crop as livestock feed..
  • Taxpayers will absorb an increasingly large share of the externalized costs of growing GM crops, as government intervention expands in the form of monitoring, tracking, regulating, and policing the growing of GM crops. 
  • In effect, taxpayers will pay three times for their GM food - first to develop (and promote(12)) it; second to monitor/regulate/police it; and thirdly, to buy it. 
  • In the event of adverse effects of GM crops on either human health (a la Walkerton) or the environment, taxpayers will again be called upon to foot the bill - unless strict liability provisions become mandatory.
  • We have seen how the growing of just one particular GM corn hybrid has imposed costs, involuntarily, on every sector of the agriculture and food system. The proprietor of the technology - Aventis - is reportedly bearing the purchase cost of buying back the year's crop of the offending grain. But that is just one part of the overall price tag for this mistake. Who is going to pay back the farmers, the elevator operators, the processors, the retail trade, or the government/taxpayers? And understand too, that this tabulation includes no consideration of health costs should StarLink actually induce allergic reactions.

    Is StarLink just one peculiar case caused by this individual hybrid? Or is it symptomatic/indicative of a pervasive problem, with foregone conclusions?

    a) Pre-harvest isolation. Is it biologically possible to reproductively isolate different cultivars of the same outcrossing species - in the absence of terminator technology? How important is reproductive isolation when the issue is not yield of the same bulk grain - e.g. corn - but batches of designer corn designed to express distinct, novel traits? How much genetic pollution can be tolerated in functional food grain fields? from functional food grain fields?

    b) Post-harvest segregation. Is it mechano-logistically possible to identity-preserve many different functional food grains on a commercial scale? To keep separate grain batches which have been produced, on a commercial scale, for high lysine or for treatment of diabetics, or for production of plastics or for any other functional traits?

    The answers are self-evident - no and no. So, what is to be done? Sooner rather than later, industry will abandon any pretense of isolation/segregation as simply unworkable, and revert to terminator type technology, to enforce pre-harvest isolation(13). Post-harvest segregation of dozens or hundreds of discrete functional food cultivars will have been shown to be equally unworkable with many thousands of independent decision-makers (retail seed trade, farmers, elevator operators), obliging yet more vertical integration to allow rigorously controlled production on a few massive subsidiary operations. The much promised benefits to farmers, for example, will be limited to those employed on the industry-contracted operations.

    The StarLink debacle has proven two fundamental truths: pollen moves, and humans are fallible. Growing crops which have been genetically modified to fulfill even a fraction of the promises which have been made - to produce drugs, plastics, industrial enzymes, plant pesticides, and nutraceuticals - can only be accomodated on a commercial scale with terminator-endowed seeds grown on industry-subsidiary farms.

    GM: a technology released prematurely

    Numerous other examples of externalities from GM crops can be cited, whether agronomic, socioeconomic, or environmental in nature (Clark, accepted; Clark, 2000). Many of these costs derive directly from the current very rudimentary state of understanding of gene function and basic plant physiology. Technology is usually viewed as the application of scientific understanding. In this case, technology is in advance of - indeed, in spite of - science, not the other way around. Brown (2000) asserted that:

    "...We must recognize that our knowledge of the processes that regulate gene incorporation and expression are in their infancy and that our capacity to manipulate the plant genome is crude. Given this current lack of understanding, it is certainly possible that the current regulatory safeguards are inadequate and may not be offering sufficient protection against inadvertent creation of health and ecological problems." and further, "Clearly the assumption that a transformed crop is exactly the sum of the original crop and the introduced gene is not acceptable. RDNA techniques are profoundly different from traditional breeding methods and are well known to cause unexpected metabolic perturbations...."

    Regarding the use of GM to modify plant metabolism to achieve industrial ends, Facchini et al. (2000) stated:

    "...these efforts to alter plant metabolic pathways....have often produced unpredictable results, primarily due to our limited understanding of the network architecture of metabolic pathways...most current models of metabolic regulation in plants are still based on individual reactions, and do not consider the integration of several pathways sharing common branch points....".

    It is therefore not surprising that one of the most widely used GM crop types - "Roundup Ready" crops tolerant to the herbicide glyphosate - has already exhibited significant, yield-limiting problems in commercial production (Coghlan, 1999; Doll, 1999; Myerson, 1997). The gene conferring glyphosate tolerance acts at the level of secondary metabolism, via the shikimic acid pathway.

    In a nutshell, GM technology has been commercialized prematurely, with a very incomplete understanding of the genetic and physiological underpinnings of the technology, and without due regard for the clear potential for externalized costs.

    Final thoughts

    The issue of labelling personifies the uniqely perverse assessment of costs and benefits in GM to date.

    Who bears the costs of labelling when the product is organic? The organic farmer, of course. The end-user of the production technology absorbs the costs. Why? Because for organic growers, identity is preserved to promote a product - "yes" organic, and to claim the premium price accorded organic products.

    Conversely, who bears the cost for labelling (identity preservation) in today's GM world? Not the end-users of the technology, those who have chosen to adopt GM crops, but rather, the non-adopters - their neighbors. Why? Because when all GM offerings pertain to agronomic rather than consumer-oriented traits, and grain is marketed in bulk, identity is preserved to assure the lack of something - e.g. "non"GMO - whether to claim GMO-free price premia or to retain certified organic status. There is no advantage, no price premia to claim, for growing GM crops.

    It does not seem imprudent to ask why it is that governments and industry continue to develop and promote products for which farmers are at pains to assert non-adoption, while at the same time damning with faint praise the development and marketing of organic products for which producers are proud to proclaim adoption.

    Now is the time for taxpayers, consumers, and environmentalists to take a position on the future - if any - of the GM industry. The federal and provincial governments currently expend an estimated $700 million of taxpayers money annually to support the biotech industry in Canada. Ontario just recently contributed yet another $9 million in taxpayers money to support an "incubator" for biotech startup companies in Toronto (Zehr, 2000). Saskatchewan and Quebec, in concert with the Royal Bank, just announced yet another $42 million (of which $28 million is taxpayers money) to create Foreign Technologies Management Inc. to "bridge the gap between a scientist's research bench and the startup of new biotech companies" (Lyons, 2000).

    Faculty time at Guelph and elsewhere is being commited to facilitate the commercialization of university research into biotechnology. Labs, infrastructural support, and graduate students are being reallocated to support ongoing, unending expansion of biotech research capability - at the expense of everything else. Graduate student defenses can now be announced as "closed" to the university community, with those allowed to participate being obliged to sign secrecy agreements, to safeguard proprietary information.

    William McDonnough once quoted someone who said "The best way to predict the future is to invent it." Is this the kind of future you want to see?

    Brown, P. 2000. The promise of plant biotechnology - the threat of genetically modified organisms. Unpublished mimeo. University of California, Davis.

    Clark, E. Ann. Genetically modified organisms and their potential impact on biodiversity. In: O. Hendrickson et al. (eds) Ecosystem Globalization: Threat to Canadian Biodiversity Environment Canada and the Canadian Forest Service, Ottawa. (accepted)

    Clark, E. Ann. 2000 (invited). Why is AgBiotech not ready for prime time? It's the process, not just the products. Presented as the Elisabeth Laird Lecture, University of Manitoba, Winnipeg. (http://www.plant.uoguelph.ca/research/homepages/eclark/laird.htm) January 2000

    Coghlan, A. 1999. Splitting headache. Monsanto's modified soya beans are cracking up in the heat. New Scientist (20 Nov. 1999)

    Colborn, Theo, Diane Dumanoski, and John Peterson Myers. 1996. Our Stolen Future. Plume/Penguin Books, New York.

    Doll, J. 1999. Glyphosate resistance in another plant. (http://www.biotech-info.net/glyphosate_resist.html)

    Facchini, P.J., K.L. Huber-Allanch, L. W. Tari. 2000. Plant aromatic L-amino acid decarboxylases: evolution, biochemistry, regulation, and metabolic engineering applications. Phytochemistry 54:121-138.

    Faeth, P., R. Repetto, K. Kroll, Qi Dai, and G. Helmers. 1991. Paying the Farm Bill: US Agricultural Policy and the Transition to Sustainable Agriculture. World Resources Institute, Washington D.C. 70 pp.

    FAO (Food and Agriculture Organization). 2000. Food Safety and Quality as Affected by Organic Farming. Agenda Item 10.1. 22nd FAO Regional Conference for Europe. Porto, Portugal, 24-28 July 2000.

    Garry, V.F., D. Schreinemachers, M.E. Harkins, and J. Griffith. 1996. Pesticide appliers, biocides, and birth defects in rural Minnesota. Environmental Health Perspectives 104(4):394-399.

    Kaufman, M. 2000. Biotech corn may be in various foods. Washington Post (19 Oct. 2000).

    Lyons, M. 2000. Sask invests in biotech venture. The Leader-Post (Regina) (20 Oct 2000)

    MacArthur, M. 2000. Triple-resistant canola weeds found in Alta. Western Producer 10 Feb 2000

    McSweeny, L. 2000. GM laws no protection, say lawyers. The Age (26 August 2000) (http://theage.com.au/news/20000826/A27644-2000Aug 25.html)

    Myerson, A. R. 1997. Seeds of discontent: cotton growers say strain cuts yields. New York Times 19 Nov 97.

    Pearce, F. 1999. Crops without profit. Britain is paying an extraordinary price for its agriculture. New Scientist 18 December 1999 (http://www.newscientist.com/ns/19991218/newsstory4.html)

    Pimentel, D. H. Acquay, M. Biltonen, P. Rice, M. Silva, J. Nelson, V. Lipner, S. Giordano, A. Horowitz, and M. D'Amore. 1992. Environmental and social costs of pesticides. In: D. Pimentel and H. Lehman (ed) The Pesticide Question. Environment, Economics, and Ethics. Chapman and Hall, New York.

    Pollack, A. 2000. Kraft recalls taco shells with bioengineered corn. New York Times 23 Sept 2000 (http://www.nytimes.com/2000/09/23/business/23FOOD.html)

    Porter, W.P., J.W. Jaeger, and I.H. Carlson. 1999. Endocrine, immune, and behavioral effects of aldicarb (carbamate), atrazine (triazine) and nitrate (fertilizer) mixtures at groundwater concentrations. Toxicology and Industrial Health 15:133-150.

    Zehr, L. 2000. Ontario set for biotech funding. Globe and Mail (17 June 2000).


    1. For an exception, see Faeth et al, 1991, who observed that "current accounting practices", which do not put a dollar value on depreciation of natural resource capital, such as soil productivity, uncontaminated groundwater, and viable wildlife habitat, "can mask a decline in wealth as an increase in income, i.e. living off your capital."

    2. According to the FAO (2000), "Virulent strains of E. coli, such as E. coli 0 157:H7, develop in the digestive tract of cattle, which is mainly fed with starchy grain as research as Cornell University has demonstrated. Cows mainly fed with hay generate less than 1% of the E. coli found in the faeces of grain-fed animals..."

    3. For example, as of early September 2000, the California legislature had already passed legislation (Assembly Bill 2622) for the governor's signature, assessing a special charge on the sale of GM rice - to cover the cost of enforcing segregation standards. Yet, as reported by McSweeny (2000) for Australia, no jurisdiction has yet come up with legislation to protect either society or the environment from adverse outcomes from GM crops

    4. Including hospitalization and outpatient treatment, lost work, accidental fatalities, deaths and reduced valuation of livestock, loss of natural enemies, cost of biocide resistance, honey bee and pollination losses, crop, fishery, and wildlife losses, groundwater contamination, and government regulation to prevent damage

    5. Unlike the other commercialized Bt crystal proteins, the Cry9C protein also has some characteristics associated with known allergens (e.g. it is heat-stable and is slow to break down in simulated digestion studies); understand too that it is the absence of these characteristics - despite no actual testing of any kind - which has been used to justify the assertion of non-allergenicity risk in all GM crops approved to date

    6. In fact, some 300 separate products have now been identified by the US

    7. this same requirement will pertain for all functional food applications of GM, the so-called "next wave" of GM inventions

    8. For this reason, Aventis has reportedly surrendered its registration for StarLink and will not be marketing it for the 2001 season, until it is approved for human use. Why these kinds of issues could not have been foreseen, and the registration withheld, is unclear

    9. Canola that is resistant to three different families of herbicides, due to cross-pollination among neighboring GM (herbicide tolerant; HT) fields over a 3-year period, e.g. Roundup Ready (glyphosate tolerant), Liberty Link (glufosinate tolerant), and Pursuit (imazethapyr-tolerant)

    10. Agronomic problems occur because HT volunteers compromise weed control options, including those of neighboring farmers. Marketing issues occur because inadvertent transfer of GM pollen to neighboring non-GM crops eliminates the GMO-free premium sought by some producers, and compromises organic status for certified organic producers - both are examples of costs imposed involuntarily to enable someone else to grow a GM crop

    11. e.g. not just a non-GM crop but a different crop entirely; e.g. wheat on a farm growing GM corn and soybean

    12. As in the multi-million dollar AAFC advert in the 13 Oct 2000 issue of Globe and Mail, and the earlier multi-million dollar advert in Canadian Living by the CFIA.

    13. It should be clearly understood that genetic pollution works both ways. Not simply would a crop grown to produce a vaccine be potentially contaminated by one grown for plastic, but a non-GM crop would just as readily be contaminated with one grown to produce a vaccine.

    14. Comments are reflective of personal opinion only and do not infer any position on the part of the University of Guelph

    ©2000 E. Ann Clark. This text is the property of the author, E. Ann Clark. It may be downloaded or reproduced in whole or in part by any member of the academic community for the purposes of discussion, debate and quotation and may be placed on web sites or on chat lines so long as this copyright notice is included. It may be reproduced on the Internet so long as no charges are levied for its use. It may not be reproduced for sale in any form anywhere without the express written permission of the owner.