Jonathan Landman, deputy managing editor of The New York Times, oversees the paper's digital media (and is a member of the Stony Brook U. School of Journalism's Professional Advisory Board). During the last week of May, he answered questions from readers in the Talk to the Newsroom feature.

His most fascinating answer came at the end, and I include it here in full:

Q. How much does the ability of online newspapers to monitor what people are actually reading affect what becomes "newsworthy"? For example, if the day's top E-mailed article is about school finance does that mean that the Times will generate more articles in the future about school finance and less about those articles that don't generate buzz? Is this different from the days of print-only journalism?

— Steve Tomasula
A. There were a lot of smart questions this week about news judgment — in particular about how the new-media environment has affected it. They were deeply informed, sophisticated questions like Mr. Tomasula’s, about serious and difficult issues we wrestle with every day. And there seemed to be a subtext: Aren’t we all going to hell in a handbasket?

I don’t think we are. In fact, I think the opposite. I think this is a remarkably exciting time for journalism, though not always for journalists. The very meaning of news and information is being redefined. Think of how Wikipedia has undermined the settled value of credentialed authority.

Credentialed authority? Oh yes, that’s us.

Rewriting rules and challenging assumptions is heady, stimulating stuff. It’s also unsettling, for us and for readers alike. We ask, “What do readers want from us that’s different from what they wanted before?” Readers ask, “Can we still trust you?”

We know we have to find new ways to tell stories, to communicate more effectively with readers, to engage citizens by helping them make information useful. When we do it people congratulate us for our spirit of adventure and push us to go farther. “It seems like you guys are throwing a lot of your value in the trash by sticking to the traditional, stovepiped newspaper model,” wrote Matt Mireles, who described himself as a 27-year-old digital native.

But readers also warn us against going too far.

Patrick Roath wrote from Somerville, Mass., to ask whether we’re comfortable with leading the news coverage on our Web site with blog posts. “I'm often surprised to see that the Times' headline story is a blog post that, while it may be up-to-date, seems to be hastily put together with little on-the-ground reporting,” Mr. Roath wrote, adding, “They often also contain a good deal of opinion and do not reflect the standards the Times has for News headline stories.”

I disagree on this point. Blogs are terrific for following a running story in real time. That’s why we use them to carry the weight of breaking news. Here’s how our Metro news blog, City Room, covered today’s fatal crane collapse in Manhattan. Read it and judge for yourself. But clearly, the fast-moving, interactive Web and the slower-paced, one-way newspaper work differently. The challenge is to be responsive without surrendering judgment; to adjust old definitions of quality to fit the new medium while staying true to our fundamental values.

Patrick Murphy, of Lionville, Pa., asks for “confirmation that you're willing to let your reporters innovate and learn from crabbed feedback” like his. (He also has problems with blogs.) The answer: Yes! Every day.

So here we are, studying the way readers use our Web site to see if we can figure out what they want. As Mr. Tomasula suggests, we can count every click. We know exactly what people are looking at online, which articles they read, how much video they consume, where they came from and where they go. Another reader, Lucas Vernon, notes correctly that some online news organizations teach reporters to put lots of keywords in the first paragraphs of their articles. This makes the articles more visible to search engines but doesn’t do much for readers who appreciate good prose. So we don't do it.

Some news executives think people should be paid more if their articles get more Web traffic. A good idea? Mr. Vernon doesn’t think so and neither do I. One of the most popular items on nytimes.com in recent memory was a slide show with pictures of celebrities in tuxedos and Oscar gowns. And fine pictures they are. Do I hear anyone arguing that they have more value to New York Times readers than this kind of reporting from the Middle East?

(Don’t lose heart, loyal Times readers! Coverage of the presidential election campaign generates tons of Web traffic and so do investigative scoops like the one that preceded Eliot Spitzer’s political demise.)

We’ll keep studying our readers closely, using the most sophisticated technology and techniques we can muster. We’ll respond by altering the architecture of our Web site, improving the way it works, providing tools to make it more open and flexible. If we discover gaps between our judgment of what is interesting and important and that of the readers who rely on us to help them sort through the noise, we will make the necessary adjustments.

After all, if people stop finding us compelling, we’re dead. Happily, that is not happening. Our Web traffic has been rising. New users are coming. In February, blogs linked to nytimes.com almost 50,000 times, more than double the next most-blogged news site, according to Nielsen BuzzMetrics. Buzz, properly understood, is good.

We won’t turn into a fanzine when we see that readers like to look at pictures of Oscar gowns. But we might add slide shows to our Movie section. No, Mr. Tomasula, we will not push Iraq off our homepage to pander to school-finance buffs! But we might start weighing the costs and benefits of building an education site.

We are lucky to have in our company a student of Web analytics named Jeffrey Graham. He’s the person who leads our effort to understand our online readers better and here’s his take on how to use reader-research wisely:

I want the site more responsive to users, because I don't think we have the same authority as an institution over Web products and technology (what gets built) as we do over what gets written (and I know that this is a blurred distinction). I think if we aren't responsive to users in what we build, we will not succeed. But I don't believe, at this point, that Web analytics should have any influence over which stories are written, how they are written, and what gets featured. It is negotiating between these two ideas that is hard. I say "at this point" because editors are always responsive to the audience — but the state of the art of analytics is not refined enough to truly represent it in a way that we can trust.

That’s it for this week. Thanks for a lot of challenging and interesting questions. Please keep them coming for my colleagues in the coming weeks.