<literal>America/New_York</literal>, consider
<programlisting>
=> SELECT '2018-03-11 02:30'::timestamptz;
timestamptz
------------------------
2018-03-11 03:30:00-04
(1 row)
</programlisting>
Because that day was a spring-forward transition date in that time zone,
there was no civil time instant 2:30AM; clocks jumped forward from 2AM
EST to 3AM EDT. <productname>PostgreSQL</productname> interprets the
given time as if it were standard time (UTC-5), which then renders as
3:30AM EDT (UTC-4).
</para>
<para>
Conversely, consider the behavior during a fall-back transition:
<programlisting>
=> SELECT '2018-11-04 01:30'::timestamptz;
timestamptz
------------------------
2018-11-04 01:30:00-05
(1 row)
</programlisting>
On that date, there were two possible interpretations of 1:30AM; there
was 1:30AM EDT, and then an hour later after clocks jumped back from
2AM EDT to 1AM EST, there was 1:30AM EST.
Again, <productname>PostgreSQL</productname> interprets the given time
as if it were standard time (UTC-5). We can force the other
interpretation by specifying daylight-savings time:
<programlisting>
=> SELECT '2018-11-04 01:30 EDT'::timestamptz;
timestamptz
------------------------
2018-11-04 01:30:00-04
(1 row)
</programlisting>
</para>
<para>
The precise rule that is applied in such cases is that an invalid
timestamp that appears to fall within a jump-forward daylight savings
transition is assigned the UTC offset that prevailed in the time zone
just before the transition, while an ambiguous timestamp that could fall
on either side of a jump-back transition is assigned the UTC offset that
prevailed just after the transition. In most time zones this is
equivalent to saying that <quote>the standard-time interpretation is
preferred when in doubt</quote>.
</para>
<para>
In all cases, the UTC offset associated with a timestamp can be
specified explicitly, using either a numeric UTC offset or a time zone
abbreviation that corresponds to a fixed UTC offset. The rule just
given applies only when it is necessary to infer a UTC offset for a time
zone in which the offset varies.
</para>
</sect1>
<sect1 id="datetime-keywords">
<title>Date/Time Key Words</title>
<para>
<xref linkend="datetime-month-table"/> shows the tokens that are
recognized as names of months.
</para>
<table id="datetime-month-table">
<title>Month Names</title>
<tgroup cols="2">
<thead>
<row>
<entry>Month</entry>
<entry>Abbreviations</entry>
</row>
</thead>
<tbody>
<row>
<entry>January</entry>
<entry>Jan</entry>
</row>
<row>
<entry>February</entry>
<entry>Feb</entry>
</row>
<row>
<entry>March</entry>
<entry>Mar</entry>
</row>
<row>
<entry>April</entry>
<entry>Apr</entry>
</row>
<row>
<entry>May</entry>
<entry></entry>
</row>
<row>
<entry>June</entry>
<entry>Jun</entry>
</row>
<row>
<entry>July</entry>
<entry>Jul</entry>
</row>
<row>
<entry>August</entry>
<entry>Aug</entry>
</row>
<row>
<entry>September</entry>
<entry>Sep, Sept</entry>
</row>
<row>
<entry>October</entry>
<entry>Oct</entry>
</row>
<row>
<entry>November</entry>
<entry>Nov</entry>
</row>
<row>
<entry>December</entry>
<entry>Dec</entry>
</row>
</tbody>
</tgroup>
</table>
<para>
<xref linkend="datetime-dow-table"/> shows the tokens that are
recognized as names of days of the week.
</para>
<table id="datetime-dow-table">
<title>Day of the Week Names</title>
<tgroup cols="2">