Re: Conceptual, Semantic, Syntactic and Lexical Design

There's an apocryphal story that the only company to apply the Foley et al.
scheme went bust (and I can name
the apocryphal source for a small donation to our SME support work ;-))

I can also name a very successful senior UK HCI person who used this stuff to good effect
(even in the face of my misplaced derision at the time) on a large Alvey demonstrator in the 1980s.

There is method in this madness.

Lexical, syntactic and semantic are machine-oriented concepts that reflect levels of abstraction
in system design.  So lexemes chunk up within a syntactic structure into semantic commands
*which may or may not* correspond to user-oriented concepts in a space defined by (select  by
ideology/faith/caprice): task, activity, domain, fact or raw indisputable intuition.

So, taking the drag to folder example, each entry of a folder with feedback is a well-formed
semantic entity all the way up from mouse moves to object transitions, but it may still have
no conceptual equivalent, and the lack of confirming feedback (folder highlight e.g,) will confirm this
to the end user.  It is thus semantically OK to drag a document to a scroll bar, but it is not generally
possible to assign a real-world (conceptual) meaning to this.  We could get deep into Fodor's Case Frames as to whether this is at the syntactic/semantic boundary, but I'm passing on that one.

So, the short answer is that the conceptual level is on Norman's user side of the gulfs of execution
and evaluation, and all the others are on the system side.  Now, there's a Mississippi or two of
water under the bridge since the Foley model, and the user side of the gulf is much less sparse than
it was in those days (see ideas from Sproull, Newman and Moran from that era on 'Conceptual' models).
These days, any self-respecting HCI warrior would have a hoard of contextual constructs where this
'conceptual'  'do-fors' (see under Lucille Ball) used to be.

In very short order, conceptual refers to the real world, everything else to the system, and the gap
between the two would be a good measure of a system's fit to its intended context of use.  One day we'll know how to expose this gap effectively and someone will get rich as a result.

Cheers
Gilbert


----- Original Message -----
From: Frank Wilson <[email protected]>
Date: Monday, November 17, 2003 5:57 pm
Subject: Re: Conceptual, Semantic, Syntactic and Lexical Design

> I agree the scheme is useful, but I always feel its expression is
> limited by
> its links to language when not all communication follows a linguistic
> metaphor. Semantic (meaning) - as you exemplify, a folder suggests
> itself to
> be a container (visual semantics), just as a 'raspberry'
> (acoustic) suggests
> an error (in my current game) - feedback! I also like the 'tring'
> (macintoshsystem X acoustic feedback for completed actions) which
> definitely sounds
> like success, and the 'bong' which could only mean failure
>

> > think so<g>), but semantic/functional is giving me trouble. If
> concepts> (e.g. a conceptual model including attributes and
> relationships) are
> > dealt with at the conceptual level and sequences (e.g. use cases
> or user
> > stories) are dealt with at the syntactic level, what does
> semantic or
> > functional design look like? The sort of thing I can think of is
> what it
> > means to drag a document to a folder. Do others agree?

    --------------------------------------------------------------
     POSTINGS (in plain text): [email protected]
           SUBSCRIPTION CHANGES:  http://lists.syntagm.co.uk
            (or send email to mailto:[email protected])
    --------------------------------------------------------------