In Nederland woedt al jaren een discussie over de rekenmodellen waarmee de natuurschade als gevolg van ammoniakuitstoot wordt berekend. Die uitstoot komt van industrie, vervoer, de natuur zelf en de mest van boerenbedrijven. De modellen die voor berekeningen worden gebruikt, laten uitkomsten zien die niet corresponderen met feitelijke metingen. Daarom worden ze steeds bijgesteld. Het gat tussen metingen en modeluitkomsten blijft echter gapen. In dit Engelstalige artikel - dat tevens bedoeld is voor de internationale gemeenschap die inmiddels geïnteresseerd is in het Nederlandse ammoniakdebat - legt chemicus en wetenschapsfilosoof Jaap Hanekamp uit dat de modellen op louter logische gronden moeten worden verworpen.
Een en ander betekent dat het Nederlandse ammoniakbeleid gebaseerd is op een oordeel van wetenschappelijke experts zonder feitelijke basis. Of het dromen zijn, is een andere vraag. Wel moet worden geconstateerd dat het beleid niet wetenschappelijk is onderbouwd. De vraag die gesteld moet worden is niet 'geeft het model een benadering van de werkelijke depositie?' (het antwoord op die vraag luidt: dat kan bij de huidige benadering onmogelijk worden vastgesteld) maar 'hoe kan de werkelijke depositie beter worden bepaald?'
Our paper “A volatile discourse” generated some discussion in the Dutch agricultural community and beyond. Our main critics - RIVM and WUR - will no doubt try to debunk our paper. And that is how it should be. This is the process by which the scientific knowledge base grows. However, there are signals that the institutions that are responsible for the ammonia research for the past 30 years have understood little of the logic of scientific debate. What we have seen so far as criticism from both the RIVM and the Wageningen University does not bode well for the coming critique to be published, no doubt, in Soil Use and Management.1 Can this debate be salvaged? Of course it can. But that requires some oversights to be rectified. We will discuss a few.
The first thing to notice is that we published only one paper fundamentally conflicting with many reports and some publications of both the RIVM and the WUR. One could easily dismiss our paper in terms of quantity. Why? Well, scientific research requires many experiments and analyses to get to grips with that part of reality under investigation. So, our case must be marginal to say the least. Right? Not so! Let me explain.
How many different empirical experiments did Pythagoras (570 to ca. 490) require in order to prove the Pythagorean theorem? Not a single one. His proof is pure logic, starting from entirely conceptual premises entailing necessary conclusions. If the premises are faultless, so must be the conclusion.
Empiricism and logic
Conversely, scientific arguments start from empirical premises and draw probabilistic conclusions prone to correction. What the empirical sciences produce are contingent propositions, not necessarily true or false: “chemical A interacts with protein X resulting in effect Y”; “the element thallium has the atomic weight of 204.38”; “the lethal dose of X for rats is Y”; “the consumption of this food adds to our health and longevity”.2
These and many other propositions generated by the empirical sciences are all (as in all) conditionally true, given various facts and evidence. None of the four propositions above are logically necessary. It is logically possible for these statements to be false, say, due to measurement errors, mistakes in experimental setups, incorrect starting materials, the limitations of available facts, and so on.
Reification
The use of mathematics thus unearthed quite the errors in this discourse. And only one such paper is needed to do so. Obviously, mistakes can be made in mathematics, but such mistakes have nothing to do with e.g. contrasting empirical results that outnumber our paper. Reducing the former to the latter is a serious, yet much-practiced, category mistake.
Another issue that should be addressed is the reification fallacy, found in many environmental debates that rely heav-ily on modelling. Reification is a widespread and classical fallacy dubbed by Alfred North Whitehead (1925) as the ‘fallacy of misplaced concreteness’. Reification is making something that is hypothetical or abstract physically real. The ammonia gap is one such reification, where models to calculate national ammonia emissions or atmospheric concentrations, as if by magic, denote the concrete reality of actual emissions or actual concentrations that do not fall in line with physical measurements.
Unquantifiability leads to begging the question
If policies are implemented buttressing on such reified models, we always end up with unaccountable and unquantifiable regulatory systems. And that is exactly what has happened here (the PAS anyone?). In the end, ammonia policies are meant to protect ecosystem’s biodiversity, but whether or not that has been successful remains a mystery, and unsurprisingly so. Anyone contesting this will immediately revert to the models I just critiqued, ending up begging the question, one of the many fallacies of argumentation we came across in this discourse.
Other fallacies we encountered are the appeal to authority and the closely related bandwagon fallacy –the appeal to popularity. The former needn’t be addressed here at length. It can never be a proxy for the content of a discourse. The latter is about the problem of disagreeing on the magnitude of some impact of ammonia emissions from agricultural sources on biodiversity or public health (via particulate matter, which is the latest hype), as it seems out of line with the majority perspective in the scientific or policy community. But that is exactly what a scientist should do.
Wholesome skepticism
No scientific results will give us definitive answers to our many questions. Many scientists, perhaps following too closely the citizen- or, worse, the policy-cheering section, developed the risky habit of insisting that their conditional truths are necessary truths. Some have gone further downhill by insisting fallaciously that their probable truths are universally true. The compelling statement “science has shown that ...” should be taken with a grain of salt, and sometimes perhaps even more than that, say, a truckload. Wholesome skepticism thus is a balancing act, as the famous chemist and philosopher Michael Polanyi showed, between orthodoxy and dissent,3 between the quietist “everybody knows that ...” and the twitchy “forget everything you know about ...”.
Scientism
This brings us to our final topic that ties in closely with the reification fallacy. And that is the false belief that human inquiry can become all-encompassing, explicitly with the aid of science. This is also known as scientism: that is the false idea that only one type of human understanding—science—is in control of the entire universe and what can be said about it. The trend analysis done by the RIVM as to show that ammonia policies were effective between 1993 and 2004 is a prime example here. Two arguments are used by the RIVM: the mentioned time interval and the ammonia emission reduction, as reported by the WUR, corresponding with the trend in said time interval.4
If the elected time interval would be special in terms of a workable ex post analysis of policy effectiveness, as the RIVM maintains, it would grant us a look into the future of ammonia concentrations. More precisely, it would, by default, give us the predictive power of policymaking and its effectiveness. Better even: I could predict tomorrow’s concentrations fairly accurately. And therewith, the first argument falls apart. Of course we cannot predict such a thing. But according to the RIVM we must. That is what their argument entails!
What is worse, the RIVM confuses epistemology with ontology. The latter gives us the why of measured ammonia concentration; that is the causal chain of events: soil chemistry, manure application, ammonia evaporation physics, atmospheric chemistry, turbulence, temperature, radiation, etc., etc. Every single step of the way will result in a cer-tain measured concentration. RIVM simply says: it’s agricultural practices and nothing else (well, for the most part). Now that is a scientistic perspective if I ever seen one. But a measurement is epistemology. It is simply the knowledge gained by measuring and recording things, such as concentrations. That doesn’t give as any insight in the why. But the RIVM knows why a priori, and thereby stumbles into a circular, question-begging, argument of epic scientistic proportions.
This whole conundrum is exacerbated by the fact that ostensible reductions in ammonia emissions as published by the WUR simply are taken for granted. We have shown that that is accepting, without any further inquiry, a non-existing, illusory, accuracy in emission numbers. That in itself undermines the whole argument, and again exposes a scientistic perspective on science and its value, accuracy, and applicability.
The ammonia discourse is a mess of poor science and even poorer argumentation in maths and logic. Overall, it lacks insight into the inner-workings of how science and logic must be partners in scientific analyses. A saner and less hyperbolic practice of science, one that is not quite so dictatorial and inflexible, one that is calmer and in less of a hurry, one that is far less sure of itself, one that has a proper appreciation of how much it doesn’t know would benefit specialist and nonspecialist alike.7
Notes
1. Our responses:
https://www.v-focus.nl/2017/01/beweringen-rivm-lijnrecht-in-tegen-conclusies-ammoniakonderzoek-hanekampcrok/
https://www.v-focus.nl/wp-content/uploads/Reactie-op-WUR-27-01-2017-Hanekamp-Crok-Briggs.pdf
2. See further: Bast, A., Hanekamp, J.C. 2017. Toxicology: What Everyone Should Know. A Book for Researchers, Consumers, Jour-nalists and Politicians. Academic Press, UK, USA.
3. Polanyi, M., 1963. The potential theory of adsorption. Authority in science has its uses and its
dangers. Science 141: 1010 – 1013.
4. See note 2.
5. See the classic http://www.thelancet.com/pdfs/journals/lancet/PIIS0140-6736(05)60233-4.pdf for a similar misleading ad hoc time frame choice and the fallacious conclusions drawn therefrom.
6. See Matt Briggs’ post http://wmbriggs.com/post/20993/ Trend Models & The Inability To Discover Cause.
7. Briggs, W., 2016. Uncertainty. The Soul of Modeling, Probability & Statistics. Springer, Switzerland.
Dit artikel afdrukken
Can this debate be salvaged? Of course it can. But that requires some oversights to be rectified. We will discuss a fewPythagoras
The first thing to notice is that we published only one paper fundamentally conflicting with many reports and some publications of both the RIVM and the WUR. One could easily dismiss our paper in terms of quantity. Why? Well, scientific research requires many experiments and analyses to get to grips with that part of reality under investigation. So, our case must be marginal to say the least. Right? Not so! Let me explain.
How many different empirical experiments did Pythagoras (570 to ca. 490) require in order to prove the Pythagorean theorem? Not a single one. His proof is pure logic, starting from entirely conceptual premises entailing necessary conclusions. If the premises are faultless, so must be the conclusion.
Empiricism and logic
Conversely, scientific arguments start from empirical premises and draw probabilistic conclusions prone to correction. What the empirical sciences produce are contingent propositions, not necessarily true or false: “chemical A interacts with protein X resulting in effect Y”; “the element thallium has the atomic weight of 204.38”; “the lethal dose of X for rats is Y”; “the consumption of this food adds to our health and longevity”.2
These and many other propositions generated by the empirical sciences are all (as in all) conditionally true, given various facts and evidence. None of the four propositions above are logically necessary. It is logically possible for these statements to be false, say, due to measurement errors, mistakes in experimental setups, incorrect starting materials, the limitations of available facts, and so on.
Our calculations are logically implied by the model’s form and assumptions. We were just the first to think to carry out the implications. As a result, we uncovered a dramatic over-certainty in national emission values. The accuracy with which these numbers are published simply does not existOur publication is in line with the Pythagorean example. We used mathematics (a form of logic) to show that the RIVM datasets of atmospheric ammonia concentrations are erroneously treated, and that the mathematical ammonia emission model of Ryden & McNeill much-used by the WUR generates substantial uncertainties, which were nev-er published. As our colleague Briggs remarked: ‘Our calculations cannot be dismissed. They are logically implied by the model’s form and assumptions. We were just the first to think to carry out the implications.’ As a result, we uncovered a dramatic over-certainty in national emission values. The accuracy with which these numbers are published simply does not exist. They are imaginary.
Reification
The use of mathematics thus unearthed quite the errors in this discourse. And only one such paper is needed to do so. Obviously, mistakes can be made in mathematics, but such mistakes have nothing to do with e.g. contrasting empirical results that outnumber our paper. Reducing the former to the latter is a serious, yet much-practiced, category mistake.
Another issue that should be addressed is the reification fallacy, found in many environmental debates that rely heav-ily on modelling. Reification is a widespread and classical fallacy dubbed by Alfred North Whitehead (1925) as the ‘fallacy of misplaced concreteness’. Reification is making something that is hypothetical or abstract physically real. The ammonia gap is one such reification, where models to calculate national ammonia emissions or atmospheric concentrations, as if by magic, denote the concrete reality of actual emissions or actual concentrations that do not fall in line with physical measurements.
Unquantifiability leads to begging the question
If policies are implemented buttressing on such reified models, we always end up with unaccountable and unquantifiable regulatory systems. And that is exactly what has happened here (the PAS anyone?). In the end, ammonia policies are meant to protect ecosystem’s biodiversity, but whether or not that has been successful remains a mystery, and unsurprisingly so. Anyone contesting this will immediately revert to the models I just critiqued, ending up begging the question, one of the many fallacies of argumentation we came across in this discourse.
Other fallacies we encountered are the appeal to authority and the closely related bandwagon fallacy –the appeal to popularity. The former needn’t be addressed here at length. It can never be a proxy for the content of a discourse. The latter is about the problem of disagreeing on the magnitude of some impact of ammonia emissions from agricultural sources on biodiversity or public health (via particulate matter, which is the latest hype), as it seems out of line with the majority perspective in the scientific or policy community. But that is exactly what a scientist should do.
Wholesome skepticism
No scientific results will give us definitive answers to our many questions. Many scientists, perhaps following too closely the citizen- or, worse, the policy-cheering section, developed the risky habit of insisting that their conditional truths are necessary truths. Some have gone further downhill by insisting fallaciously that their probable truths are universally true. The compelling statement “science has shown that ...” should be taken with a grain of salt, and sometimes perhaps even more than that, say, a truckload. Wholesome skepticism thus is a balancing act, as the famous chemist and philosopher Michael Polanyi showed, between orthodoxy and dissent,3 between the quietist “everybody knows that ...” and the twitchy “forget everything you know about ...”.
Scientism
This brings us to our final topic that ties in closely with the reification fallacy. And that is the false belief that human inquiry can become all-encompassing, explicitly with the aid of science. This is also known as scientism: that is the false idea that only one type of human understanding—science—is in control of the entire universe and what can be said about it. The trend analysis done by the RIVM as to show that ammonia policies were effective between 1993 and 2004 is a prime example here. Two arguments are used by the RIVM: the mentioned time interval and the ammonia emission reduction, as reported by the WUR, corresponding with the trend in said time interval.4
Of course we cannot predict such a thing. But according to the RIVM we must. That is what their argument entails!Apart from the fact that both arguments are blatantly false, the scientistic (reductionist) perspective evinced here baggers belief. The selected time interval facilitates an ad hoc argument that is inflated to the level of actual proof of policy effectiveness. However, 1993 – 2004 is a completely random timeframe (that, however, fits nicely with the looked for outcome).5 Why? We could have started in 1980 to look at atmospheric ammonia concentrations, or 1970. The riposte would be obvious: the RIVM started sampling in 1993, and not before that date. But that is completely irrelevant. I would like to know ammonia concentrations before that date. There is nothing special about 1993, or 1980, 1970 of 1433 for that matter. And there is nothing special about 2004 as well.6
If the elected time interval would be special in terms of a workable ex post analysis of policy effectiveness, as the RIVM maintains, it would grant us a look into the future of ammonia concentrations. More precisely, it would, by default, give us the predictive power of policymaking and its effectiveness. Better even: I could predict tomorrow’s concentrations fairly accurately. And therewith, the first argument falls apart. Of course we cannot predict such a thing. But according to the RIVM we must. That is what their argument entails!
The ammonia discourse is a mess of poor science and even poorer argumentation in maths and logic. Overall, it lacks insight into the inner-workings of how science and logic must be partners in scientific analysesEpic scientism
What is worse, the RIVM confuses epistemology with ontology. The latter gives us the why of measured ammonia concentration; that is the causal chain of events: soil chemistry, manure application, ammonia evaporation physics, atmospheric chemistry, turbulence, temperature, radiation, etc., etc. Every single step of the way will result in a cer-tain measured concentration. RIVM simply says: it’s agricultural practices and nothing else (well, for the most part). Now that is a scientistic perspective if I ever seen one. But a measurement is epistemology. It is simply the knowledge gained by measuring and recording things, such as concentrations. That doesn’t give as any insight in the why. But the RIVM knows why a priori, and thereby stumbles into a circular, question-begging, argument of epic scientistic proportions.
This whole conundrum is exacerbated by the fact that ostensible reductions in ammonia emissions as published by the WUR simply are taken for granted. We have shown that that is accepting, without any further inquiry, a non-existing, illusory, accuracy in emission numbers. That in itself undermines the whole argument, and again exposes a scientistic perspective on science and its value, accuracy, and applicability.
The ammonia discourse is a mess of poor science and even poorer argumentation in maths and logic. Overall, it lacks insight into the inner-workings of how science and logic must be partners in scientific analyses. A saner and less hyperbolic practice of science, one that is not quite so dictatorial and inflexible, one that is calmer and in less of a hurry, one that is far less sure of itself, one that has a proper appreciation of how much it doesn’t know would benefit specialist and nonspecialist alike.7
Notes
1. Our responses:
https://www.v-focus.nl/2017/01/beweringen-rivm-lijnrecht-in-tegen-conclusies-ammoniakonderzoek-hanekampcrok/
https://www.v-focus.nl/wp-content/uploads/Reactie-op-WUR-27-01-2017-Hanekamp-Crok-Briggs.pdf
2. See further: Bast, A., Hanekamp, J.C. 2017. Toxicology: What Everyone Should Know. A Book for Researchers, Consumers, Jour-nalists and Politicians. Academic Press, UK, USA.
3. Polanyi, M., 1963. The potential theory of adsorption. Authority in science has its uses and its
dangers. Science 141: 1010 – 1013.
4. See note 2.
5. See the classic http://www.thelancet.com/pdfs/journals/lancet/PIIS0140-6736(05)60233-4.pdf for a similar misleading ad hoc time frame choice and the fallacious conclusions drawn therefrom.
6. See Matt Briggs’ post http://wmbriggs.com/post/20993/ Trend Models & The Inability To Discover Cause.
7. Briggs, W., 2016. Uncertainty. The Soul of Modeling, Probability & Statistics. Springer, Switzerland.
Nog 3
Je hebt 0 van de 3 kado-artikelen gelezen.
Op 5 mei krijg je nieuwe kado-artikelen.
Op 5 mei krijg je nieuwe kado-artikelen.
Als betalend lid lees je zoveel artikelen als je wilt, én je steunt Foodlog
Lees ook
Goed werk Jaap. Waarom hebben jullie zo lang werk gehad om dit op papier te krijgen daar je jaaaaren geladen al overtuigd was van hetgeen je nu opgeschreven hebt. Wat nu met alle kosten die de veehouderij gemaakt heeft t.a.v. ammoniak reductie. Maar wat nog erger is de impact op de humane gezondheid. We hebben inzichten die er niet om liegen, de indirecte schade die het inbrengen van mest in de grond teweegbrengt aan ons spijsverteringsysteem is enorm. Het toenemen van de problemen met ons auto immuunsysteem, het heeft er alle schijn van dat het niet op orde zijn van de bodem hier debet aan is. En het niet op orde zijn van de bodem heeft ook een relatie met, hoe we omgaan met organisch materiaal, Niet onderwerken. Het bodemleven moet zijn natuurlijke werk doen. Wat nu als de schade, aangericht aan de humane gezondheid, in beeld komen, wie is dan verantwoordelijk?
Arnold, Jaap zegt slechts zegt dat de modellen geen afbeelding van de werkelijkheid zijn. Jij leest er jouw ding in en - in de termen die Jaap gebruikt -reïficeert ook weer. Daar waarschuwt Jaap tegen: laten we toch eens leren met onzekerheid om te gaan en toch verstandig te zijn.
Kijk nu toch wat een onzin we hier weer lezen over het onderzoek waar Jaap het hierboven over heeft. Daar doet hij 0 uitspraak over. Dat is een boer die zijn gelijk haalt met een oneigenlijk argument. Let's get real, eindelijk!
Arnold, en lees dit: Annie Schreijer-Pierik is tenminste van de wetenschap. Geloof jij het?
Dick Veerman,
Goede reacties, maar Annie is van het CDA en de Bonnetjes de ze weg gooit en tevens boerin (geweest) en boeren denken een generatie vooruit lees ik nu. Prachtig het gemanipuleer en het gaan geloven in hetgeen goed uitkomt.
Je reactie in 2 ook goed, de boer in het artikel schrijft een aantal dingen waar ik het mee eens ben zoals dat mestinjectie nadelen heeft op de bodem en de bodem weer op de gezondheid van het vee.
De conclusie die de boer in het artikel trekt dat het beleid in het geheel niet of zelfs averechts werkt heeft Prof Jaap C. Hanekamp inderdaad in het geheel niet gezegd.
Ik merk het vaak bij boeren (ook in de familie) het is niet zozeer het gaan geloven in het eigen gelijk, het is gaan geloven in hetgeen goed uitkomt (meestal in de euro's) en daar medestanders voor zoeken.
Terug naar het hoofdartikel,
Prof Jaap C. Hanekamp heeft het wat mij betreft bij het rechte eind of beleidsmakers nu ander beleid gaan voorstellen en of dat de wetenschappers die aanvankelijk het beleid mee opgetuigd hebben over hun eigen schaduw (of ego) heen kunnen stappen is weer een andere vraag.
Ammoniakbeleid is een (goedkoop) sturingsinstrument geworden m.b.t. ruimtelijke ordening, zie Brabant.
Beste mensen, dat ammoniak zorgt voor mega schade in de natuur, dat is toch allang duidelijk en wetenschappelijk gezien een feit. Dat je in plaats van mest injecteren (of wassers plaatsen) beter de veestapel kunt halveren of, beter nog, tot een kwart terug brengen, ja, dat is zeker waar.
Ontkennen dat ammoniak schadelijk is voor mens en natuur, dat is een achterhoede gevecht dat je niet zult winnen. Verspilde tijd en ontkenning van veel onderzoek.