If people on Earth want to talk to aliens, we may have to change our tune - or at least the way we broadcast it. And we may need a crowd to help figure out what to do.

A team of three scientists recently wrote a paper suggesting that humans need to change protocols in order to have a message to possible aliens understood. Previous attempts at contacting aliens, known as messaging extra-terrestrial intelligence (METI), might be heard someday, but there is little hope that aliens will be able to decipher what the message is about (though they might see that it is artificial).

The first time humans tried signaling aliens was in 1974, from the Arecibo radio telescope in Puerto Rico. The message was aimed at a globular cluster called M13, 25,000 light years away. (An answer, if one comes, will come back in 50,000 years).

That message consisted first of the numbers one through 10, then the atomic numbers of basic elements that make up DNA. After that, it had encoded the formulas for the sugars and bases.

If an alien continued to decode the message it would see the number of nucleotides in DNA, a graphic of the double helix, a figure of a human with the average size and the approximate human population at the time. Further number-crunching would reveal a graphic of the solar system and the Arecibo telescope.

Other attempts were made in 1999, 2001 and 2003, radio messages sent from a tracking facility Yevpatoriya in Ukraine. The 1999 and 2003 messages were digital transmissions designed to be visible against background noise. The 2001 attempt was called the Teen Age Message and used musical tones to make the signal stand out better over interstellar distances. The signals were sent to stars within a few light years.

The last try was in 2008, and was a digital signal sent to Gliese 581c, a planet detected around Gliese 581, 20 light years away. It consisted of text messages and photos submitted by the public.

Dimitra Atri, the lead author on the paper, writes that the problem is that all of these messages, especially the last one, depend on being picked up by some creature that shares human senses. For example, a photographic message would make no sense to a creature without eyes.

To better ensure that a message is understood, there needs to be a protocol. That protocol can be tested by involving the public in developing it, the paper says. One way to do it is via a web site that allows people to encrypt a message, while other users try to figure out what it is. Testing a message protocol this way can help determine how culturally biased, for example, a message is. It can also help the people transmitting the message see where improvement is needed.

For example, students in the United States may construct a set of messages according to the [messaging extra terrestrial intelligence] protocol, which are then exchanged with students in China, the paper says. The differences in cultures between the respective student groups will likely be reflected in the message composition, so not all messages may be successfully communicated. In this case, students in the United States may discover that their particular ideologies that they believed to be obvious were somewhat lost in translation when deciphered by their international colleagues. Repeating this process with diverse groups of people can help refine the protocol to be more universally understandable.

Even though there are no aliens (yet) to test the protocol on, a message that can be deciphered across cultural boundaries is a good start, Atri says in the paper. It is often said that SETI is a search for ourselves, and as we develop a message that we would send to unknown listeners, we will come to an even deeper appreciation of our diversity as humans, the paper says.