Apparently we didn't know how to read a thermometer in any of the decades of the last century. They all had to be 'adjusted'.
Then we found out that we didn't know how to read thermometers in the 00's. And everything had to be 'readjusted'.
In three years we will find out that we didn't know how to read a thermometer in the 10's either. And all the readings will have to be re-readjusted.
In actual fact, the numbers are being constantly changed. The readings for pre-1960 typically go down, and after 1960 they typically go up.
The abrupt change at 1940 is the result of two things. Necessary but still discretionary adjustments to ocean temperatures. And a need to smooth out warming/cooling bump that wrecked the temperature to CO2 correlation.
Here is part of an email chain released in Climategate-
From: Tom Wigley <wigley@ucar.edu>
To: Phil Jones <p.jones@uea.ac.uk>
Subject: 1940s
Date: Sun, 27 Sep 2009 23:25:38 -0600
Cc: Ben Santer <santer1@llnl.gov>
It would be good to remove at least part of the 1940s blip, but we are still left with "why the blip"
Edit- remember these are just changes from 2008. No one knew that the adjustments were going to be made and the old data tossed out, or at least made unavailable. All the most reasonable and necessary adjustments had already been made by the turn of the millennium.