I wish I were a lawyer so I could figure this out. I tried searching for info on the patents and came up with this article that goes into more detail on them:
FOSS Patents: Apple requests U.S. preliminary injunction against the Samsung Galaxy Nexus based on four high-power patents
One of the things it mentions is that the patent involving siri could be a big problem for Google since Google's main deal is the search engine. THAT DOESNT MAKE ANY FREAKING SENSE!!! GOOGLE HAS HAD THEIR SEARCH ENGINE AROUND FOREVER!!! SIRI JUST CAME OUT!
At least part of this may be the patent system. The US Patent system is so overloaded, the primary research for granting a new patent is a search among existing patents. While it's entirely true that "prior art" (eg, the "invention" already in use) will invalidate a patent, there's a process gap here. If your patent is granted, the presumption from then on is that you have a 20-year monopoly (from the date of application) on that invention.
I'm not a lawyer, but I am an Engineer who's been involved in patents from both sides. Year ago, I was the technical liaison for my company's legal team, going up against a gigantic company that decided to sue us. More recently, I have written a number of patents.
In this specific case, it's very possible, perhaps even likely, that Google has been doing the same thing as SIRI since before SIRI. There's no guarantee -- I haven't read the patents in question, or done any other analysis, just suggesting what might be. Google has had voice-directed search in Android nearly since the beginning, and it has improved. Could be that they just didn't file any patents on this, thinking it well represented in the prior art (in theory, the patent applicant has to disclose all knowledge of applicable prior art as part of the patent application, but I know of a large number of patents filed on things "obvious to one skilled in the art" with no such disclosure, which would have been rejected given that disclosure).
And so Apple, even acting on their best behavior, comes up with the same basic idea, and, not knowing (again, in theory) about Google's work, files and receives patents. They shouldn't have been, but again, the PTO is not omniscient, and Apple is at best being as not-knowing as they can legally document. One of the problems with software patents in particular is that there are so many of these, it's virtually impossible for even a good patent search to turn up everything (and trust me, I have done many patent searches, I'm very good at it, and still don't always find everything). And that's assuming the best behavior.
There have been worse. When software patents were first allowed, in the early 80s, there was kind of a gold rush. The whole concept wasn't even for a software patent -- there was an important case in which the patent couldn't be denied simply because the hardware system used a software component -- prior to this, inventions were routinely rejected due to a software component. After all, software is already covered under copyright, eh?
So you had companies like IBM, just submitting all kinds of nonsense patents. For example, they were granted a 1984 patent on cut & paste between text buffers. It's impossible to imagine that any software engineer (and hard to imagine any engineer) in those days had not already used a text editor that did this. I had used RMS's original TECO Emacs, and Gosling's C Emacs, back in college, before this. They not only did this, but the very same keyboard sequence used to claim my company (Commodore) and product (Amiga) violated this patent would have worked on the Emacs I used back in '79. But even then, the primary basis for "is it patentable" was drawn from the pool of patents.
The system really needs an extended period of public review. I'm not holding my breath.
And I'm also not saying Apple's wrong here, either.. again, I haven't read the patent. I'm just trying to offer an explanation as to why these things keep happening. The patent system itself is to blame -- it's highly flawed, and it will continue to grant patents on things that have long existed.
The other thing that also involves court is the very nature of what a patent is. Patents don't cover ideas, they cover very specific implementations of ideas. Or they're supposed to. So it's quite possible that Apple's patent doesn't apply to Google just because Google does what appears to be the same thing, but the underlying architecture of the software is somewhat different. That would actually keep Apple's patent intact, but absolve Google of any infringement. The problem is that, when you write the application, you're generally limiting the scope quite a bit, to have it pass muster. Once you're granted the patent, you might get aggressive.
As we know, Apple's goal isn't licensing, it's spreading FUD and randomly throwing monkey wrenches into the Android world. So they may consider it a win just to delay a Google product to market by six months. And to get all that press, because the average reader doesn't have a clue about how the patent system actually functions.