Index ← 2411 CFJ 2412 2413 → text
==============================  CFJ 2412  ==============================

    If Rule A says that a person SHALL perform an action by a specific
    mechanism, but Rule B says a person SHALL NOT perform an action by
    the mechanism, and no other mechanism for performing the action is
    specified, then the person CAN perform the action via the mechanism
    (though perhaps breaking Rule B).

========================================================================

Caller:                                 G.

Judge:                                  ehird
Judgement:                              TRUE

========================================================================

History:

Called by G.:                           13 Mar 2009 17:31:06 GMT
Assigned to ehird:                      13 Mar 2009 17:36:20 GMT
Judged TRUE by ehird:                   14 Mar 2009 16:34:27 GMT

========================================================================

Caller's Arguments:

We've accepted that SHALL implies CAN (e.g. CFJ 2120) when no other
CAN mechanism is specified.  But if one rule contains a SHALL, does
that enable the CAN to work even if another rule says SHALL NOT?

Or is the net effect of the rules that matters (e.g. SHALL -> CAN
only if the net effect of the rules says SHALL, so the relative
precedence of the SHALL versus the SHALL NOT matters)?

The issue here is that CAN and SHALL work differently when conflicts
arise:

1.  A conflicting CAN and CANNOT results in a net single "Rules say
    CAN or CANNOT" based on precedence;
2.  A conflicting SHALL and SHALL NOT does not create a conflict
    based on impossibility, so both exist in the rules, merely
    creating a difficulty in obeying both.

So how does that translate for SHALL->CAN?

========================================================================

Gratuitous Arguments by G.:

While we've found that SHALL -> CAN, we haven't found that SHALL NOT ->
CANNOT.  In fact, accepting that SHALL NOT -> CANNOT would probably
break a lot of things.

========================================================================

Gratuitous Arguments by omd:

SHALL and CAN are supposed to be orthogonal, and it's
perfectly reasonable to require someone to perform an action by a
certain mechanism without actually allowing them to (e.g., you SHALL
deregister by submitting a Cantus Cygneus rather than by
announcement).  The special case of a simple 'X SHALL do Y by
 was established in CFJs 1765 and 1890 to imply the
mechanism, but I'd argue that's just an archaic linguistic shortcut,
not a necessary side-effect of the SHALL.

========================================================================

Gratuitous Arguments by G.:

Sorry, you're right; but the CFJ statement is indeed for cases where
it's "X SHALL do Y by mechanism Z"; so as long as we're assuming that the
archaic shortcut is functional, it's relevant.  -G.

========================================================================

Gratuitous Arguments by omd:

But I argue it's _just_ a linguistic shortcut: not some odd attribute
of 'CAN-ness' attached to the obligation, but an expansion to apply
within the rule's text before anything else is considered.

'SHALL by announcement' --> 'CAN by announcement, and SHALL'

'SHALL NOT by announcement' --> well, I guess this has never been
tested.  Is it 'CAN by announcement, but SHALL NOT' or just 'SHALL NOT
use the announcement mechanism to ...'?

Either way the answer is TRUE because one or both rules say CAN and
neither say CANNOT.

========================================================================

Gratuitous Arguments by G.:

That's reasonable.

But you do leave out the third possibility for 'SHALL NOT by announcement',
which is 'CANNOT by announcement and SHALL NOT'.  Not that it's a
particularly likely construction, I agree, but its unlikeliness is as much
a historical accident as the original shortcut.

========================================================================

Judge ehird's Arguments:

"X SHALL Y by Z" is fairly obvious shorthand for "X CAN Y by Z and X
SHALL Y using the mechanism Z". The two things are separate.

And "SHALL NOT" does not in any way imply CANNOT. There are actions we
CAN perform, but SHALL NOT. The SHALL NOT only cancels the SHALL; not
the CAN. (In fact, the SHALL might take precedence of it; I'm assuming
SHALL NOT power > SHALL).

========================================================================