Original research

Most funeral poem lists publish work they probably cannot publish

Families searching for a funeral reading land on a small set of pages. We checked what those pages actually reproduce. Of the thirteen we could analyze, eleven published work that is in copyright or of disputed status, and one stated the public-domain status of anything at all. The full list of pages is published below so the count can be checked.

Published August 7, 2026 · Methodology and limitations below

11 of 13

pages reproduced at least one identifying line of a work that is in copyright or of disputed status

8 of 13

reproduced at least 70 percent of the lines of at least one such work

1 of 13

stated the public-domain status of any poem it published

0 of 13

showed a visible permission or licence statement beside a reproduced modern poem

Why this matters to a family

A funeral program is a reproduction. When a family copies a poem from a website into a program and prints two hundred copies, they have republished it, and the fact that the printing was small, unpaid, and grief-stricken does not change that.

In practice, families almost never face consequences. The people who do are the ones printing at scale: funeral homes, print shops, template sellers, and stationery businesses. At least one frequently requested modern poem has a rights holder with a documented history of pursuing unlicensed use.

The deeper problem is that families cannot tell the difference. A poem published online with no author credit looks free. Very often it is not free; it is simply uncredited, and the missing credit is what makes it look safe.

What was being reproduced

Ordered by how many of the thirteen pages carried at least one identifying line of each work. The second figure counts pages reproducing at least seventy percent of its lines.

Do Not Stand at My Grave and Weep

9 of 13 · 5 in full

Attributed to Mary Elizabeth Frye, c. 1932

Disputed. Authorship was contested for decades and the copyright position has never been conclusively settled.

She Is Gone (also published as He Is Gone)

7 of 13 · 6 in full

David Harkins, 1981

In copyright. Widely misattributed to an anonymous author, which is likely why it is reproduced so freely.

Afterglow

6 of 13 · 6 in full

Helen Lowrie Marshall, 1958

In copyright. Frequently published with no author credit at all.

Miss Me But Let Me Go

5 of 13 · 2 in full

Anonymous, disputed

Disputed. No settled authorship, which leaves its status unresolved rather than clear.

Funeral Blues (Stop all the clocks)

4 of 13 · 3 in full

W. H. Auden, 1938

In copyright in the United States and most other countries. Demand rose sharply after its use in film.

When Great Trees Fall

3 of 13 · 1 in full

Maya Angelou, 1990

In copyright.

To Those Whom I Love

3 of 13 · 3 in full

Isla Paschal Richardson

In copyright.

The Dash

2 of 13 · 2 in full

Linda Ellis, 1996

In copyright and actively licensed. The rights holder has historically pursued unlicensed reproductions, including by individuals and small businesses.

Turn Again to Life

2 of 13 · 1 in full

Mary Lee Hall

Uncertain. Attribution and first publication are both unclear.

God Saw You Getting Tired

1 of 13 · 1 in full

Anonymous, disputed

Disputed authorship and unresolved status.

The Broken Chain

0 of 13 · 0 in full

Ron Tranmer

In copyright, but not counted on any page in this sample. The lines most often used to detect it are traditional floating stanzas that also appear inside other, differently credited poems.

The pages analysed

Published in full so that anyone can repeat the count. Listed alphabetically, not in ranking order. The complete dataset, including per-page line-coverage figures and every search engine attempt, is available as JSON. Verbatim lines of the poems are withheld from it, because republishing them is the practice this study measures.

  • amandalouisefuneralservices.co.uk
  • dignityfunerals.co.uk
  • eulogyexpert.com
  • familytreeceremonies.co.uk
  • gupton-jones.edu
  • humanist.org.uk
  • humanists.uk
  • panmacmillan.com
  • poets.org
  • stoneletters.com
  • tharpfuneralhome.com
  • theinspiredfuneral.com
  • wvfuneralboard.com

Two further domains were found but could not be analysed, and are excluded from every figure on this page:

  • dignitymemorial.com Returned HTTP 403 to automated access. Publishing practice unknown.
  • reddit.com Returned only 198 characters of server-rendered text, so the page content could not be verified.

One page did it properly

wvfuneralboard.com It marks public-domain works as such, and beside Maya Angelou it prints 'Copyrighted. Available in Angelou's published collections. Do not reproduce — description only.' For the Harkins poem it prints no text at all, only a pointer to where the poem can be found. It is the one page in the sample that states copyright status, and an earlier version of our own detection rule would have wrongly scored it as a reproducer because the title of several of these works is also their opening line.

How this was done

On August 7, 2026 we collected the top ten organic results for "funeral poems" and "funeral poems and readings" from Brave Search and deduplicated by registrable domain, producing 15 unique publishers. 2 could not be analysed, leaving 13 pages.

Brave Search was used because every other engine we tried refused automated access that day: Google, DuckDuckGo, Startpage, Bing. We did not attempt to defeat any anti-bot measure. This matters when reading the result, because the choice of engine determines the sample.

Each page was fetched and reduced to plain text. We then measured how many lines of eleven works that are in copyright or of disputed status appeared on it, counting a work as reproduced in full at seventy percent line coverage or above. A bare title, or a note telling readers where a poem can be found, is not counted as reproduction. That distinction matters: for five of these eleven works the title is also the opening line, so a simpler rule would score a page that merely names a poem as though it had published it.

The full-text figure is not sensitive to where that line is drawn. It holds at eight of thirteen at every coverage threshold from sixty to ninety percent.

Separately we recorded whether the page stated public-domain status anywhere, and whether any permission or licence statement appeared beside a reproduced modern poem.

What this does not establish

  • This is a count from one named, dated sample, not a property of 'the top-ranking pages' in general. Search results are personalised and change daily, so a repeat run will return a different domain list.
  • Google, DuckDuckGo, Startpage and Bing all refused automated access on the run date, so Brave Search was used. A different engine indexes different pages, and this sample came back heavily weighted toward United Kingdom publishers even though the requests originated from a United States address. That weighting directly affects which poems appear.
  • Thirteen pages is a small sample. A single page moves any proportion by roughly eight percentage points.
  • Reproduction was detected by matching identifying lines and measuring coverage across all of a work's lines. Bare title mentions and pointers telling readers where to find a poem are not counted as reproduction. Partial quotation of a few lines may be defensible in some jurisdictions.
  • The eleven works are a fixed list, so the study measures how often those particular poems are reproduced, not how common copyright-risky publishing is overall. One page in the sample reproduced none of the eleven yet published other plainly in-copyright poems in full.
  • A publisher may hold a licence that is not visible on the page. The zero figure measures disclosure, not conduct, and one of the pages reproducing a modern poem in full is that poem's own trade publisher.
  • Copyright status was assigned from author death dates and publication history. No page was checked against any rights holder's actual licence records.
  • Two domains in the sample, humanist.org.uk and humanists.uk, are separate domains belonging to the same organisation. Deduplicating by organisation instead of by domain would reduce the sample to twelve.
  • This is a description of publishing practice, not legal advice, and it is not an accusation against any individual publisher.

What we did about it

We publish our own list of funeral readings, so the honest response was to hold it to the standard this study measures. Every poem and passage on it carries its publication year, a link to the source text, and a plain statement of why it is free to print. The same page names the frequently requested readings that are not free to print, including several found repeatedly in this study.

See the verified readings

Questions

Is it illegal to print a poem in a funeral program?
It depends entirely on the poem. Reproducing a work that is in the public domain is unrestricted. Reproducing a work still in copyright without permission is an infringement, even in a small, non-commercial print run handed out at a service. Reading a work aloud at a private gathering is treated differently from reproducing its text.
How can a family check a poem before printing it?
Find the year of first publication and the author's death date. In the United States, anything first published in 1930 or earlier is in the public domain. Elsewhere the common rule is the author's life plus seventy years. If the poem circulates online with no author credit, treat that as a warning sign rather than as evidence that it is free, because uncredited circulation is exactly how modern poems end up presumed to be anonymous.
What should a funeral home do about this?
Most exposure comes from template libraries and printed programs that reproduce whatever families request. A short internal list of verified public-domain readings, offered to families as the default, removes almost all of the risk and takes very little effort to maintain.
Can we cite this study?
Yes. The methodology, the sample size, and the limitations are stated on this page so that the finding can be checked. Please link to this page so readers can see how the sample was constructed and what it does not establish.

Citation and corrections

Cite as: TributeReady, “Most funeral poem lists publish work they probably cannot publish,” August 7, 2026, sample of 13 pages from Brave Search, tributeready.org/funeral-poem-copyright. This page describes publishing practice in aggregate and does not accuse any individual publisher of infringement. If you publish one of the pages analyzed and hold a licence, or if you believe a work is categorized incorrectly, write to care@tributeready.org and we will correct the page.