Learning and Mislearning 77HYVs, use fertilizer and transplant. They said they would do so only if their water
supply and other inputs were assured. This was agreed. Special efforts were made
with the bank, and with seed and fertilizer supply, to ensure that farmers could
obtain the necessary inputs. The subminor had a continuous flow all through the
season with a flow of 4 to 5 cusecs to a lower area which was itself never short of
water. Farmers could always see water flowing past their outlets, even when they
were not receiving it.
To the best of my knowledge and belief, none of the controls with which the
comparisons were made received the same special treatment of negotiation, guar-
anteed water supply, or arrangements for credit and input supply.
In the light of these procedures and other evidence let us now look at each of
the four findings.
Acceptance by cultivators. No indication is given of how the ‘general acceptance’
was assessed, nor whether this included the two trial subchaks which were rejected,
but it is scarcely surprising if the Koliary outlet cultivators accepted assurances of
a guaranteed water supply, credit and inputs. Indeed, the benefits from these could
override other inconveniences if they were a necessary price to pay. Had there been
negative attitudes to subdivision and rotation, they could have been obscured by
welcome for unprecedented access to credit and other services. In sum, this finding
cannot be said to have been established.
Time taken for irrigating the entire outlet command was 5 to 14 days as against
20 to 45 days for normal outlets without suhchaks. The figures cited in the finding
are not those of the report. The comparison of 5 to 14 days with 20 to 45 days
looks decisive and impressive. However: if the averages for all the chaks are taken,
the average for trials was 15 days and for controls 17 days, hardly a decisive differ-
ence; the 5 to 14 figure for trials is reached by accepting the two better performing
trial chaks, which were reported to take 5 and 14 days respectively, and rejecting
the trial chak on HBP which was reported to have taken 25 days. No control chaks
were rejected; the 20–45 days figure for controls is more difficult to understand.
The report gives a range of 6 to 46. Four of the controls, taking 6, 6, 8 and 10 days
respectively were said to have had heavy rainfall which shortened their irrigation.
This rainfall was not reported to have affected any of the trials. But even if these
four control chaks are eliminated on the grounds of heavy rainfall which they
alone received, the controls remaining took 7, 8, 16, 17, 22, 27 and 46 days respec-
tively, giving a range of 7 to 46, not 20 to 45.
Thus, instead of 5 to 14 for trials and 20 to 45 for controls, the correct com-
parison would appear to be 5 to 25 for trials and 6 or 7 to 46 for controls. The
figures officially quoted as ‘findings’ are difficult to explain.
A more plausible interpretation is that time taken to irrigate the entire outlet
command was a function of position on the main system and of the water supply.
This is suggested, for example, by comparing the two trials with head locations on
the MRP with the MRP control which had a comparable head location. The per-
formance was closely similar, as follows: