Although Marx uses 'productive' in more than one sense, Angelus is right that 'productive labour' _mainly_ has a specific meaning in Capital that has everything to do with whether it creates surplus value or not and nothing to do with whether it is necessary to social reproduction in a broader sense. But I'm with Julio in saying, 'so what?' Definitions should serve analysis, not the other way around. It was worth being clear about the distinction, but not worth dragging the discussion into Great Quotation-slinging.
Julio's also right that if you think Capital is a finished, coherent conceptual system, you haven't read it. And the worst thing about so much Marxology is that many people who can quote the text itself back-to-front know very little about the early-to-mid-19th-century political economy that is its context and even so much of its content. If your economic methodology is to rip selections from the text 150 years out of that context, you're often anachronistically defending the concerns and framing of Georgian and Victorian political economy rather than any particularly radical criticism of economics past or present. The 'labour theory of value' is a case in point - Marx's theory points the way out of the labour theory, but thanks to the conservatism of exegesis-based economics, which freezes analysis at an earlier stage of its development, it's become synonymous with it! Close the hermeneutic circle, people!
Finally, I can't resist a little Marxology myself: contra Carrol, the fact that Capital is subtitled 'a critique of political economy' does not mean that it is not also a positive contribution to political economy, or that we should not engage in economics. Kant's book is called the 'critique of pure reason' not because he wanted to rule out of court all a priori judgments, but to inquire as to what makes them possible. Likewise, Capital is an inquiry as to what makes political economy possible, and proceeds by way of a lot of reflection on what makes economic phenomena appear as such, as a relatively autonomous social sphere open to a semi-autonomous social science - and conversely, the limits to their autonomy resulting from the embeddedness of the economy in a historically-developing social whole.
Mike Beggs