Disclaimer - not historian just a hobbyist.

It is a myth, at least as far as Ellis Island is concerned. Individuals coming to the US might change their name to help with assimilation process. It could also have been done to signify a fresh start to a new life in the states.

Some immigrants also changed their names to make them more "american." By that I mean that their name could be more easily pronounced.

As the second link suggests, many name change stories are just that, family stories. They provided a flawed connection to the family's initial entrance to the states.

Source - http://www.smithsonianmag.com/smart-news/ellis-island-isnt-blame-your-familys-name-change-180953832/

http://www.uscis.gov/history-and-genealogy/genealogy/genealogy-notebook/immigrant-name-changes

In the film, the Immigration official didn't deliberately change Vito's name- he misheard the Italian translator and marked his name down as "Corleone." It was a bureaucratic mix-up that Vito never corrected.

Hello everyone,

In this thread, there have been a large number of responses which break our rules concerning personal anecdotes, only relating personal, family stories. While they're sometimes quite interesting, they're unverifiable, impossible to cross-reference, and not of much use without more context. This comment explains the reasoning behind this rule. Please, before you attempt answer the question, keep in mind our rules concerning in-depth and comprehensive responses. Answers that do not meet the standards we ask for will be removed.

Additionally, it is unfair to the OP to further derail this thread with off topic conversation, so if anyone has further questions or concerns, I would ask that they be directed to modmail, or a META thread. Thank you!

In addition to the fact that there weren't actually Ellis Island name changes, surnames based on geographical origin are common in Italy anyway. Italians have more surnames than any other ethnicity in the world--despite the fact that regular people only really started using them after the middle ages. Most of them are based on personal characteristics (ie: Russo means red, so like, someone with red hair), who their father is (ie: pretty much any surname beginning with "Di"), what they did for a living (ie: Palmieri means "palmer" or priest), OR where they were from (Like Napolitano, Romano, etc.).

Interestingly, the very common last name Esposito possibly indicates that someone's ancestor was an orphan (espositi were orphans, derived from the same latin word as our word "exposed").

I know we're not doing personal anecdotes here, but just for example, my last name is also the name of a very small village in Italy. My family didn't move here from there, they moved from another town, but it is possible that they lived there originally--although the name is both descriptive and possibly indicative of an occupation.

I'm not as familiar with the naming conventions of other ethnicities, but I'd imagine that geographical surnames are common in most.

I feel the prevailing view that no errors were made at Ellis Island is not correct. I actually feel stronger than that, but I'll leave it as 'not correct.' A moderator has told me, essentially, to put up or shut up. So, here's to "putting up." Quoting from the Smithsonian magazine citation:

“No names were changed at Ellis Island because no names were taken at Ellis Island.” Instead, inspectors only checked the people passing through the island against the records of the ship on which they were said to arrive. If the name was misspelled, it was done so on the ship’s manifest documents when a person bought their ticket in Europe. (Some immigration clerks on Ellis Island even helped correct these mistakes.) Regardless, these spellings didn’t typically follow people to their new lives in America.^1 [my emphasis]

Note the closing equivocation, despite the opening broad claim to perfection. The cited USCIS (US Citizenship and Immigration Services) "Immigrant Name Changes" website makes this similar claim:

[The manifests] were created abroad, beginning close to the immigrant's home when the immigrant purchased his ticket. It is unlikely that anyone at the local steamship office was unable to communicate with this man. His name was most likely recorded with a high degree of accuracy at that time. It is true that immigrant names were mangled in the process. The first ticket clerk may have misspelled the name (assuming there was a "correct spelling"--a big assumption). If the immigrant made several connections in his journey, several records might be created at each juncture. Every transcription of his information afforded an opportunity to misspell or alter his name. Thus the more direct the immigrant's route to his destination, the less likely his name changed in any way.^2 [my emphasis]

In my view, a mangled name is a changed name. Now lets look at the New York Public Library's effort entitled "Why Your Family Name Was Not Changed at Ellis Island (and One That Was)". It starts

Between 1892 and 1954, over twelve million people entered the United States through the immigration inspection station at Ellis Island, a small island located in the upper bay off the New Jersey coast. There is a myth that persists in the field of genealogy, or more accurately, in family lore, that family names were changed there. They were not. Numerous blogs, essays, and books have proven this. Yet the myth persists; a story in a recent issue of The New Yorker suggests that it happened. This post will explore how and why names were not changed. It will then tell the story of Frank Woodhull, an almost unique example of someone whose name was changed,...^3

If you are left-brained, the NY Public Library has provided our disproof by example.^4 "Frank Woodhull, an almost unique example" [emphasis added] is all we need to throw a flag on the claim of Ellis Island perfection. QED. Note the caveats in all the citations.

But most of us hereabouts are not left-brained, so let's try to bound the issue, and Google "data entry error rate." The first return^5 gives "the average benchmark of a 1% error rate in manual data entry." The third return,^6 from Panko and Shidler of the University of Hawaii, cites a study by Mattson & Baars [1992] "Typing study with secretaries and clerks. Nonsense words. Per nonsense word." This study gives an error rate of 7.4%. European last names might be considered more or less nonsense words, but Panko and Shidler say "the error rate for more complex logic errors is about 5%".^6 With 12 million people we have 12 million last names. 1% of errors in 12 million is 120,000; 5% is 600,000; 7.4% is 888,000. All non-zero numbers.

Behind the Ancestry.com paywall on webpage http://search.ancestry.com/cgi-bin/sse.dll?indiv=try&db=nypl&h=4028173829 is the arrival manifest of SS Aquitania at New York, 28 Sep 1923. My grandfather and the ticket selling clerk did their work properly (See the discussions above.) Someone at Ellis Island, not of the cruise line, changed out a vowel.

^1 Smithsonian Magazine. http://www.smithsonianmag.com/smart-news/ellis-island-isnt-blame-your-familys-name-change-180953832/#pABcoDkROUjz4Hk2.99 Accessed 16 July 2015

^2 USCIS. http://www.uscis.gov/history-and-genealogy/genealogy/genealogy-notebook/immigrant-name-changes Accessed 16 July 2015

^3 New York Public Library. http://www.nypl.org/blog/2013/07/02/name-changes-ellis-island Accessed 16 July 2015

^4 "If we can find one member of the specified set for which the example of a member of the set for which the specified properties do not hold is called a counterexample of the statement. Stating a counterexample of a conditional statement will thus disprove the statement. Note that here by 'disprove it' I mean 'prove it to be false.' " Disproof by Counterexample. http://www.personal.kent.edu/~rmuhamma/Philosophy/Logic/ProofTheory/DisproofByCounterexample.htm Accessed 16 July 2015

^5 When Good Info Goes Bad: The Real Cost of Human Data Errors – Part 1 of 2. http://ungerboeck.com/blog/when-good-info-goes-bad-the-real-cost-of-human-data-errors-part-1-of-2 Accessed 16 July 2015

^6 Basic Error Rates. http://panko.shidler.hawaii.edu/HumanErr/Basic.htm. Accessed 16 July 2015