The fact is the internet is an ocean of limitless depth from which defamation will rise from here to infinity.

This is part two of a two-part look at defamation reform. Read part one here.

Defamation law, like much of the law, is a compromise between conflicting rights -- specifically, the right to free speech and the right to a good reputation. As we all learn the first time we express our honest opinion about bedtime, free speech is never unqualified.

Australia's defamation law struck its peculiar balance heavily in favour of reputation in a time when defamation took only two forms: publication in print (libel) and speech to a live audience (slander). It did not contemplate forms of communication like radio, let alone the internet.

Generations later, we're still tinkering with this old law, while the gap between technological and legal development widens exponentially.

That is the context for the next round of (what will be) hotly contested defamation law reform, the "stage two" discussion paper just released by New South Wales Attorney-General Mark Speakman. It poses this question: what should be the defamation liability of internet intermediaries?

This question urgently needs an answer. The courts have been tying themselves in knots for years trying to apply the analogue law to a digital world. There is also a very important commercial context because litigation trends follow the "deep pocket" theory, and the deepest pockets nowadays are owned by the tech giants. So the marketplace is demanding clear answers.

Back in 2002, the High Court bravely decided that the internet was no different from any other medium of publication or broadcast. This technology-neutral stance has been maintained by courts and legislators ever since but it's been, well, problematic.

It's not a problem when online publication is merely the digital reproduction by a mainstream publisher (i.e. a media company) of material it would otherwise publish in print. However, it gets tricky in a number of current contexts, which will be joined by others not-yet contemplated. These include hosting someone else's content online; hosting a page on which others can post comments or reviews; "liking", reposting or in some other form advancing the visibility of another's publication (e.g. a tweet); generating and posting search results with or without "snippets" of the original content (Google); aggregating and republishing material online.

That's a random selection, illustrating the immediate problem of categorisation. Nevertheless, regulators and legislators are still keen to fall into exactly that trap.

The courts have so far determined that a person or organisation hosting a web page or forum on which others can post comments can be liable for defamatory material in those comments if they fail to act (i.e. take them down) once put on notice of their existence. That is, they're not a publisher of the defamation initially ... but they can become so if they leave the material up when they could and should have taken it down.

That idea got a turbocharging last year when the NSW courts held that media companies could be liable for defamatory comments posted on their Facebook pages, whether or not they know about them or actively moderate the pages. That's going to the High Court.

Google has been considered liable for a while now with respect to its own search results. Last year a Victorian court also found it liable for defamatory content in hyperlinked pages within its search results.

Facebook, on the other hand, hasn't gone down for anything published on its site. Its argument is that it doesn't publish anything -- rather it provides blank pages for its users to write on.

This is obviously a mess. The discussion paper asks many questions about where the dividing lines should be drawn, and what non-litigious mechanisms might be created or enhanced to slow the clogging of the courts with petty Facebook disputes.

Most obviously, since the tech giants have absolute control of their own products -- they like to say otherwise, but their decision to finally scrub Donald Trump from the internet last year put an end to that pretence -- much attention can and should be focused on take-down processes to avoid court.

That won't please the lawyers and legal industry bodies, who have a solid record of arguing against any defamation reform that might reduce the volume of litigation (maybe that's just a coincidence). But see if you can find a defamation plaintiff who ever had a good time suing.

As was said by everyone after the Geoffrey Rush suit, nobody really wins in a defamation case.

Really we're having the wrong conversation here. The starting point for reform should be a simple acknowledgment: that the internet is an ocean of limitless depth which will generate defamation from here to infinity.

A legal system response designed for a world of physical publication and providing an arbitrary amount of money as its only means of repairing reputational loss is hopelessly outgunned by the ubiquity of what exists in the modern digital world. And it always will be, no matter how much we tinker at its edges.

Is defamation law salvageable? Write to [email protected] Please include your full name to be considered for publication in Crikey's Your Say section.