The rule on whether jurors may discuss a case before closing arguments is almost always a matter of statewide law and not a decision left to a particular judge in a particular case.
There is good reason to believe that prohibiting pre-closing deliberation by jurors tends to make ultimate jury verdicts more accurate for reasons discussed below.
The question about juror questions seems to be based upon the inaccurate assumption that jury questions are only allowed if collectively approved by the jury.
The Rationale For Preventing Pre-Deliberation Discussions Among Jurors
The Traditional Rule Described Is Not Universal
First of all, prohibiting jurors from deliberating or discussing a case prior to the close of evidence and closing arguments is not universal practice. Colorado, for example, is one of the minority of reformist states that allow jurors to deliberate at any time that they are all present together in a case, after the panel is chosen. Arizona was one of the first states to experiment with this innovation.
But, there is a rational reason for the older majority rule.
Discouraging Pre-Judgment Of The Case
This is first, to discourage jurors from reaching conclusions before all of the evidence is in, because people have a natural tendency to place more emphasis on the evidence that they hear first and to jump to conclusions before hearing all of the evidence, such as the defendant's evidence that comes last.
Talking to other jurors consolidates and solidifies opinions and that makes it harder to change opinions when new evidence confronts an original opinion. Groups cling more tightly to an initial consensus than individuals. In contrast, individual jurors often change their minds during the course of a trial.
Secondly, it is to maximize the independence with which jurors reach their conclusions at the close of evidence.
This matters because, empirically, the ultimate decision of a jury panel is overwhelmingly consistent with the initial pre-deliberation inclination of the median juror on the issue of guilt or innocence.
The ultimate verdict of a jury is almost completely foreordained by the initial impressions of the jurors when they start their deliberations, even though it doesn't feel like that because jury deliberations cause people who are in the minority in an initial straw poll on the verdict to change their minds based upon what the other jurors have to say about 90% of the time.
Twelve Angry Men (a drama and movie in which a holdout juror convinces all of the other jurors to change their mind on their verdict from their first straw poll) is an extremely outlier case that almost never happens in real life, especially in non-death penalty cases.
So getting the first straw poll right is much more important than getting deliberations right since those have very predictable results. Deliberations almost never matter unless the jury has a 50-50 split in their initial straw poll and a lot of cases where there is a 50-50 split in the initial straw poll will produce a hung jury.
Statistics on jury decisions
To put some flesh on the numbers, the best empirical estimates of the overall accuracy of jury panels in making guilt-innocence determinations is about 90%. (Estimates on false convictions range from about 2.3% to 7% for all serious felonies including convictions arising from plea bargains in a 2007 study, a 14% acquittal rate was found in stranger rape cases where DNA evidence can lead to acquittals following jury trials, and roughly 40-47% of death sentences are overturned in the appellate and post-conviction review process.)
Note that about 90% of criminal cases, overwhelmingly those with more clear cut guilt, result in plea bargains, so only about 1% of criminal defendants who are convicted are wrongfully convicted as a result of an inaccurate jury verdict.
This accuracy rate somewhat underestimates how hard it is to be that accurate. It has been shown statistically (in the same article supporting the majority rule of jury verdict outcomes linked above), that about 68% of cases are pretty much guaranteed to produce convictions regardless of the makeup of the jury or the process used, while about 14% of cases that produce acquittals are sufficiently clear that the makeup of the jury or the process used doesn't matter (in a large, representative Sarasota, Florida sample). Thus, only about 18% of cases presented to juries are actually close cases and on those cases juries get the right result only a little better than 50% of the time. It is reassuring that about 95%+ of cases that are prosecuted have predictable outcomes in the system either through plea bargaining or correct jury verdicts, but not very comforting that about 2%+ of criminal prosecutions, and about one in five criminal jury trials, involve such unclear facts that their outcome at trial is basically a coin flip.
In general, as a matter of mathematics, it is universally true (subject to assumptions that are present in this situation) that when something is more than 50% likely to happen in each trial as you would like it to happen, you are more likely to get a good result if you maximize the number of independent determinations that are made. In contrast, if something is less than 50% likely to happen as you would like, you maximize your odds of a desired result with as few independent trials as possible. The leading article explaining this idea is called "How To Gamble If You Must."
In a jury context, in real life, any given independent juror who ends up on a jury poll is more than 50% likely to get the right answer in an initial straw poll the vast majority of the time.
So, the more independent jurors you have, the more likely it is that a majority of them will get the right answer. And, the fewer independent jurors you have, the more likely it is that a majority of them will make a mistake. So, by preventing pre-judgment deliberation, you increase the number of independent evaluations of the evidence that are made by the jurors and increase the probability of a correct result on the merits.
The Case For The Modern Rule Allowing Pre-Closing Deliberation
The counterargument for the modern rule is that an overall mistake on guilt or innocence flows from multiple earlier mistakes in interpreting particular items of evidence, and that the ultimate conclusion flows from a lot of mini-determinations. And, if someone makes a wrong mini-determination early on and can't be swayed from it once deliberations come along, this can lead to the jury being wrongfully hung.
In contrast, if jurors deliberate on each item of evidence as it comes along as a complete group, the likelihood of getting any particular sub-issue wrong is minimized, and getting sub-issues right consistently may be more important to reaching a correct decision than maintaining independent points of view throughout the trial.
Should Juror Questions Be Allowed?
The traditional position is that jurors can't ask questions at all. To the best of my knowledge, juror question asking is a very recent phenomenon (probably late 1980s or later).
The modern position, followed in a minority of states, allows jurors to ask questions if they are not objectionable under the rules of evidence that apply to lawyers. Some jurisdictions allow this in both civil and criminal cases, while others allow it only in civil cases. I honestly don't know what the rule in Colorado is on juror questions in criminal cases, because my practice is exclusively non-criminal. The state of the law on jury questioning in criminal cases in the federal courts is summarized in a recent 6th Circuit opinion:
Juror questioning, we have explained, “should be a rare practice,” but
“the balance of risks to benefits is more likely to weigh in favor of
juror questions in complex cases.”
The concern about allowing juror questions in criminal cases is that the burden of proof beyond a reasonable doubt is on the prosecution and it can lose simply by failing to meet that burden of proof.
An answer to a juror question could provide additional evidence that turns a case where it seems likely that a defendant committed a crime but there has not been proof beyond a reasonable doubt, into one where there is proof beyond a reasonable doubt, unfairly and unconstitutionally disfavoring the criminal defendant.
In a civil case where the preponderance of the evidence burden of proof makes the balance equal between the parties, and where the focus is more on getting a right result between two people who are equal before the state as opposed to protecting people from state overreach, the same considerations about aiding one side in meeting a burden of proof with jury questions don't apply to the same extent.
The argument for the modern position is that psychology research shows that people make more accurate decisions when they can ask questions on points about which they are unclear. The lawyers and judge in the room may think that a point made by a witness was perfectly clear, but some of the jurors may not understand the word that someone used and come to the wrong conclusion. But, lawyers can't clarify the facts so as to assist the jurors in making the right decision if the jurors can't ask questions that illustrate what the jurors aren't understanding of the presentation before them.
Deliberation As Related To Juror Questions
A related issue is asking questions. Obviously if the jury wants to
ask witnesses questions, they may need to talk among themselves to
decide what question to put to the witness, so the instruction would
seem to directly conflict with basic juror rights, such as the right
to question witnesses.
In my experience when juror questions are allowed, any juror who wants to ask a question may do so unilaterally without reaching the agreement of the other jurors, on the theory that even one juror who doesn't understand something about the testimony or has unanswered questions when facts are available to them, is one too many. Usually, a few jurors ask lots of questions, and many jurors ask few or no questions.
So, it isn't clear how a prohibition on jurors deliberating prior to the close of evidence would change the practice of jurors asking questions.
Perhaps there are places where jurors are allowed to ask questions only if they agree on those questions, but I am not aware of any such jurisdiction. In every jurisdiction that I am aware of that allows questions, individual jurors can ask them, and questions are not asked by jurors only if no juror has any questions for a witness.
Also, even if jury questions did require the approval of the entire jury, my experience with how juries interact is that most juries would routinely approve almost every question of almost every individual juror anyway, so I don't know that it would make much of a difference.