A very good new book on military intelligence in the Civil War is Peter G. Tsouras' Major General George H. Sharpe and the Creation of American Military Intelligence in the Civil War. Sharpe was the director of the Army of the Potomac's Bureau of Military Information (BMI), which was created in 1863. Tsouras makes several points regarding prisoner interrogation.
At the time, soldiers were given little or no instruction of what they could/should say if captured. The modern day rule about restricting information to name, rank, and serial number were not yet inplace. Even if the PW merely gave their regiment, division, and corps, that was very useful information for a competent intelligence element in building order of battle information. It is useful to know that Longstreet's Corps is to your front. It may be even more useful to know that Longstreet isn't there, and hasn't been there for two weeks [because he's been moved to Tennessee to support Bragg and is preparing for the Battle of Chickamauga].
Deserters were motivated to cooperate with their captors, and desertion from the Army of Northern Virginia was rampant. By deciding to desert, they had abrogated their former loyalties. Cooperative deserters, whose information was confirmed from other sources, were paroled by the BMI and moved North once they had sworn the oath of allegiance. The BMI attempted to make sure that this opportunity was made known to soldiers in the Army of Northern Virginia. Similarly, PWs who did not want to be paroled back South, or exchanged, were useful sources of information. It was better to be housed, feed, clothed, and safe in a Federal PW pen than starving and at risk with Bobby Lee's army. By 1864, large numbers of Confederate PWs were refusing exchange.
Torture is the refuge of the stupid and the incompetent. A judicious application of pain will produce a confession to any crime or matter that the interrogator wants a confession to. My teaching point is the good, Catholic fathers of the Dominican Order, who caused tens of thousands of "witches" to be burned alive during the Inquisition. The "witches" had all confessed, because being burned alive was preferable to what the good fathers were doing to them to save their souls. None of that made the confessions in the slightest degree true.
From a military intelligence standpoint, torture is counter-productive. A man who knows nothing will create out of whole cloth what he thinks you want to hear simply to make the pain go away. The intelligence officer then chases his tail, at a great waste of time and effort, following and trying to confirm the resulting false leads. It also raises the ethical problem for the commander of possibly populating his intelligence organization with sadists, although that particular pathology was probably ill understood at the time of the Civil War.
Regards,
Don Dixon